98th 98hCnrsCongress COMMITTEE PRINT I S.98-221PRT.

THE EMPLOYEE RETIREMENT INCOME SECURITY ACT OF 1974: THE FIRST DECADE

AN INFORMATION PAPER

PREPARED FOR USE BY THE SPECIAL COMMITTEE ON AGING UNITED STATES SENATE

AUGUST 1984

This document has been printed for information purposes. It does not offer findings or recommendations by this committee

U.S. GOVERNMENT PRINTING OFFICE 37-569 0 WASHINGTON: 1984

For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 SPECIAL COMMITTEE ON AGING JOHN HEINZ, Pennsylvania, Chairman PETE V. DOMENICI, New Mexico JOHN GLENN, Ohio CHARLES H. PERCY, Illinois LAWTON CHILES, Florida NANCY LANDON KASSEBAUM, Kansas JOHN MELCHER, Montana WILLIAM S. COHEN, Maine DAVID PRYOR, Arkansas LARRY PRESSLER, South Dakota BILL BRADLEY, New Jersey CHARLES E. GRASSLEY, Iowa QUENTIN N. BURDICK, North Dakota PETE WILSON, California CHRISTOPHER J. DODD, Connecticut JOHN W. WARNER, Virginia J. BENNETT JOHNSTON, Louisiana DANIEL J. EVANS, Washington JEFF BINGAMAN, New Mexico JOHN C. RoTHER, Staff Director and Chief Counsel STEPHEN MCCONNELL, Deputy Staff Director DIANE LIFSEY, Minority Staff Director RoIN L. KROPF, Chief Clerk

(II) PREFACE

Ten years ago, the Employee Retirement Income Security Act of 1974 [ERISA] was passed by the Congress (Public Law 93-406), with the protection of the pension and welfare benefit rights of workers and their beneficiaries as its goal. Pensions had developed many years earlier, of course, but the enactment of ERISA marked the growing importance of private pensions in the retirement income equation. The Senate Special Committee on Aging, as the oversight com- mittee charged with the broad mandate to assess the adequacy and security of all retirement income policy and programs is pleased to mark the anniversary of ERISA with this comprehensive assess- ment of its accomplishments and the job that still remains. The first pension plans developed in the latter part of the 19th century in this country. Still, by 1910 there were no more than 100 private plans in the United States, including a few established by labor unions. Spurred by a burst of pension plan formation follow- ing the Second World War, the private pension system now covers approximately 50 million nonagricultural wage and salary workers, about 51 percent of the work force. As the first cohort of employees to enjoy this expanding pension coverage grows older, and the pri- vate pension system matures, increasing numbers of Americans expect to receive some form of employer-provided retirement bene- fit. In order to insure that these expectations would be fulfilled, ERISA prescribes minimum standards of participation, vesting, and plan funding, reporting and disclosure requirements, and standards of fiduciary responsibility. Together with the creation of plan ter- mination insurance, these provisions represented a significant step toward guaranteeing the pension promise made to employees par- ticipating in private pension plans. This year marks the 10th anniversary of the enactment of ERISA, but it also marks the transformation of tax-qualified pri- vate pensions into vehicles designed to provide retirement income for employees at all income levels, especially those who might oth- erwise have to rely solely on Social Security for their retirement income. It is therefore an ideal time to applaud its successes as well to assess its failures. The purpose of this committee print is fourfold: To recall the circumstances surrounding ERISA's passage and the goals which Congress hoped to achieve; to examine our ex- perience with ERISA enforcement and subsequent enactments; to examine the impact which ERISA has had on the universe of pri- vate pension and employer-provided retirement plans; and to raise some of the major questions concerning pension policy which remain today. The print is a compendium of seven independent essays, each providing a different perspective on ERISA.

(III) We would like to gratefully acknowledge the support of the Sun Co. in the preparation of this document. The first chapter, written by Michael Gordon, former pension counsel to Senator Jacob Javits and currently a partner in the law firm of Mittelman and Gordon, gives an insider's view of the causes of and circumstances surrounding the enactment of ERISA. He re- calls the crucial role that Senator Javits, often styled the "father of ERISA," played in bringing the need for pension legislation to the attention of Congress and in shaping the form that ERISA eventu- ally assumed. In the second essay, Dr. Thomas Woodruff, former Executive Di- rector of the President's Commission on Pension Policy, and cur- rently the executive director of the Commission on College Retire- ment, examines in detail the goals which Congress sought to achieve through the enactment of ERISA. He considers both what ERISA did and did not attempt to accomplish, focusing on the sub- sequent experience of plan participants. In chapter 3, authored by Dr. Dan McGill, chairman and re- search director of the Pension Research Council, Wharton School of the University of Pennsylvania, legislation passed since the enact- ment of ERISA is summarized and its most important elements dis- cussed. The chapter folows this legislation chronologically, draw- ing out of the enormous complexity of recent pension legislation the trends in congressional interest and activity as legislators have sought to fill the gaps left by ERISA, and expand its scope where possible. In the fourth chapter, Dr. Beverly Klimkowsky, a public policy consultant specializing in retirement income issues, evaluates the implementation of ERISA regulation in the executive branch. Dr. Klimkowsky, whose dissertation entitled "Assessing ERISA Amid Changing Expectations," reviewed in detail the problems encoun- tered by the Department of Labor, Internal Revenue Service, and Pension Benefit Guaranty Corporation in translating the mandate of ERISA into effective agency regulation and enforcement, here distills the experiences and reaction of members of the administra- tion and the pension community to the massive task faced by these agencies when ERISA became law. She concludes with the observa- tion that, given the incredible obstacles to effective implementation created by the size and scope of ERISA, the fact that the three agencies are able to administer it at all is a significant accomplish- ment. The fifth essay, prepared by Dallas Salisbury, president of the Employee Benefit Research Institute, considers the impact that ERISA has had on the growth and composition of the private pen- sion universe. The development of a cohesive pension policy has to some degree been hampered by the lack of complete, current infor- mation; this chapter examines the most recent survey data and draws some tentative conclusions. Many factors affect pension plan formation, including the economy, employee preferences, and tax incentives. Although it is difficult to separate the impact of ERISA from these other factors, identifiable trends have emerged which the author notes and discusses. In the sixth chapter, Donald Grubbs, an actuary employed by George B. Buck Consulting Actuaries, Inc., and a frequent source of technical analysis at congressional hearings, presents in a very concise manner the broad spectrum of pension issues still actively debated 10 years after ERISA's passage. His paper runs the gamut of current pension issues and controversy, and describes the argu- ments presented by advocates and critics of proposals concerning ERISA minimum standards, the burdens of Federal regulation, and emerging issues-many of which were never anticipated by the drafters of ERISA. The seventh and final chapter differs in its style and objective from the preceding six. Written by Alicia Munnell, senior vice president and director of research for the Federal Reserve Bank of Boston, the paper diverges from the largely retrospective nature of the first six essays to present the current pension policy debates. This chapter presents one view of some of the major themes in pen- sion and related policies and should serve as a departure point for a continuing dialog between Congress and the pension community about the implications of ERISA for tax, employment, investment, and retirement income policies. Private pensions are a critically important part of a larger retire- ment income system. On the tenth anniversary of ERISA's enact- ment, it is fitting that we assess whether we have made good our commitment to make sure that the pension promise is kept for the millions of Americans who will one day retire from active employ- ment. The authors of this committee print review the accomplish- ments of the last decade, setting the stage for the committee's prin- cipal celebration of this anniversary: the ERISA 10th Anniversary Conference on September 11, an opportunity for the Congress to discuss our agenda for retirement income in the next decade. JOHN HEINZ, Chairman. JOHN GLENN, Ranking Minority Member. CONTENTS

Page P reface ...... III Chapter 1. Overview: Why was ERISA enacted?: A . Introduction ...... 1 B. A capsule history of private pension plan growth...... 2 C. Prolog to ERISA-the Welfare and Pension Plans Disclosure Act and the President's Cabinet Committee Report of 1965 ...... 6 D. Period of gestation-congressional investigation, studies, and bills, 1965-72...... 10 E. The drive for enactment ...... 15 F. Epilog-reflections on the spirit of ERISA...... 24 Chapter 2. The goals of ERISA and the impact of ERISA on plan participants: A . Pensions by the 1970's...... 26 B . The goals of E R ISA ...... 28 C. What ERISA did not attempt ...... 30 D. What ERISA accomplished for participants...... 31 E. What ERISA failed to do for participants...... 32 F. Are ERISA's goals in conflict with Federal tax policy?...... 42 G . Conclusion ...... 43 H . Bibliography...... 44 Chapter 3. Post-ERISA legislation: A . Introduction ...... 45 B. Tax Reduction Act of 1975...... 45 C. Tax Reform Act of 1976...... 47 D . Revenue A ct of 1978 ...... 48 E. Multiemployer Pension Plan Amendments Act of 1980...... 52 F. Economic Recovery Tax Act of 1981 [ERTA] ...... 62 G. Tax Equity and Fiscal Responsibility Act of 1982 [TEFRA]...... 65 H. Tax Reform Act of 1984...... 70 I. Concluding observations...... 78 Chapter 4. Implementation of ERISA in the executive branch: A . Introduction ...... 80 B. Principles of implementation...... 81 C. Overview: Implementation of ERISA ...... 83 D. Stage 1: Startup phase ...... 84 E. Stage 2: "Settled dust" of normal implementation...... 95 F. Stage 3: Mature implementation...... 99 G . Con clusion ...... 101 H . B ibliography...... 102 I. Selected references on ERISA legislative oversight: 1974-84 ...... 102 Chapter 5. What impact has ERISA had on different types of pension plans?: A . In trodu ction ...... 107 B. What is a pension plan under ERISA? ...... 108 C. What was the impact of ERISA?...... 110 D. Development of employer-sponsored retirement plans...... 113 E. Employer-sponsored individual effort...... 119 F. Development of coverage for employer-sponsored retirement plans...... 120 G. Individual retirement accounts [IRA's]...... 124 H. How has growth affected savings?...... 125 I. Con clusion ...... 126 Chapter 6. Continuing policy issues: A . In troduction ...... 128 B . Coverage ...... 129 C. Accelerated vesting requirements...... 130 VIII Page Chapter 6. Continuing policy issues-Continued D . Portability ...... 132 E. Pension accruals after normal retirement age...... 135 F. Benefits for spouses and survivors...... 138 G. Integration of private pension plans with Social Security...... 142 H. Protecting pensions from inflation...... 145 I. Plan termination insurance ...... 147 J. Termination of pension plan with asset reversion...... 152 K. Investment of pension plan assets...... 156 L. Federal administration of pension policy ...... 159 M. Federal employee retirement plans...... 161 N. Pension plans for State and local government employees...... 169 0 . C onclusion ...... 171 Chapter 7. ERISA: Is it consistent with other national goals?: A . Introduction ...... 172 B. ERISA and Federal tax policy...... 174 C. ERISA and employment policy...... 181 D. ERISA and retirement income policy ...... 190 E. ERISA and capital formation...... 200 F . Conclusion ...... --.... -...... 205 THE EMPLOYEE RETIREMENT INCOME SECURITY ACT OF 1974: THE FIRST DECADE

Chapter 1 OVERVIEW: WHY WAS ERISA ENACTED? (Prepared by Michael S. Gordon, Esq.,* Partner in the Law Firm of Mittelman & Gordon, Washington, DC) A. INTRODUCTION The reasons for ERISA's enactment are inseparable from the dy- namic stages of the legislative process that preceded enactment. While the significant factors can be isolated from that process and discussed as part of impersonal socioeconomic trends, such as the growing labor mobility that took place after World War II, the drastic expansion of retirement lifestyles, or the demographic shifts in the aging population, et cetera, that approach fails to give a real sense of the specific circumstances and events that played such a crucial role in determining the need for, and the nature and scope of, pension reform legislation. In point of fact, concerns over the fairness and soundness of private retirement programs had arisen long before any serious drive for reform had gotten underway, and even after these concerns had been thoroughly cataloged and ana- lyzed, there was very little impact on public opinion-at least not enough to create the political motivations that are essential for the success of any major institutional reform. Accordingly, it is the emergence of pension reform objectives into political consciousness which deserves the closest scrutiny and is the main focus of this chapter. At its core, the particular legislative response we call ERISA is the net result of those proposals which successfully ran the gauntlet of interest group pressures, legislative and executive branch rivalries, as well as public opinion, and how it got that way will also tell us a great deal about why ERISA was enacted. An explanatory note concerning the scope of this chapter is in order. The recounting of the legislative struggle to enact ERISA is a lengthy one and is more suitable for a book rather than the brief overview intended here. For this reason, and because the period prior to 1973 is not covered by the official legislative history of

Copyright 1984 by Michael S. Gordon; reprinted with the author's permission. Mr. Gordon specializes in labor law and employee benefits. During the period 1970-75. he served under an appointment by former Senator Jacob K. Javits (R-N.Y.) as minority counsel for pensions, U.S. Senate Labor and Public Welfare Committee, and participated in the drafting of ERISA. He also served as the Department of Labor's legal counsel to President Kennedy's Committee on Corporate Pension Funds, etc., which is discussed more fully in this chapter. (1) ERISA put together by the then Senate Labor and Public Welfare Committee with the assistance of the Department of Labor,' a delib- erate decision has been made to concentrate on the pre-1973 period, where the roots of ERISA really lie, and to let the material concern- ing the final stages preceding ERISA's enactment, including the Senate-House conference deliberations of 1974, go for another time. Perhaps at some future date similar attention can be paid to the post- 1972 period, when the leading concepts of private pension reform were refined into their precise and final statutory formulations, but for now the focus is on how those concepts arose and were translated into legislative proposals. B. A CAPSULE HISTORY OF PRIVATE PENSION PLAN GROWTH The first public policy factor cited in the "Findings and Declara- tion of Policy" section of ERISA, 2 as the basis for imposing nation- al regulatory standards on employee benefit plans is their "rapid and substantial" growth in recent years. Prior to World War II, however, that growth was anything but rapid and substantial. The earliest efforts to provide retirement benefits occurred during the late 19th century in railroads and allied industries, the first private pension plan being established by the American Ex- press Co. in 1875. By 1910, about 50 percent of railroad employees were members of pension plans. By the late 1920's, the number in- creased to about 80 percent. Although there were less than 10 pri- vate pension plans in the country by the turn of the century, the total had reached nearly 100 by 1910. These included plans estab- lished by a handful of labor unions, the first union plan generally recognized as that established by the Pattern Makers in 1900. Be- tween 1910 and 1920, about 150 new plans were established in such industries as public utilities, iron and steel, oil, banking, and man- ufacturing, including establishment of the landmark Sears, Roe- buck & Co. profit-sharing plan in 1916, which consisted entirely of Sears' own stock. After World War I, the revenue legislation of first 1921, then 1926, brought tax exemption status to both stock-bonus and profit- sharing plans and pension trusts respectively. By 1925, there were about 400 private pension plans in operation covering about 4 mil- lion employees, but about one-third of these were employed by four of the largest corporations: U.S. Steel, the Pennsylvania Railroad, the New York Central Railroad and American Telephone & Tele- graph Co.3 By the onset of the Great Depression of the 1930's, many private pension plans were bankrupt. In addition, many of them were inse- cure, inflexible and discriminatory.4 Insecurity was traced to weak financing, actuarial unsoundness, inadequate legal safeguards and vague administrative procedures. Inflexibility arose out of excessive

1See the three-volume Legislative History of the Employee Retirement Income Security Act of 1974, prepared by the Subcommittee on Labor, U.S. Senate Committee on Labor and Public Wel- fare, 94th Cong., 2d Sess. (1976). 2 ERISA, § 2, 29 U.S.C. § 1001. 3 S. Rep. No. 92-634, 92d Cong. 2d Sess. 7 (1972). 'M. Latimer, Industrial Pension Systems, vol. II (New York National Industry Conference Board, 1932), at 902, 922-930. age and service requirements. The plans were discriminatory be- cause the employees in the upper echelons benefited at the expense of those in the rank and file.5 The coverage of private pension pro- grams in the 1930's was just above 10 percent of all wage earners and only between 5 and 10 percent of the numbers covered were estimated as eventually qualifying.6 The inadequacies of private pensions were, for the most part, submerged in the cataclysm of the Great Depression and the wiping out of millions of dollars in personal savings. Yet, the psy- chological environment for the future impressive growth of private pensions was created by these adverse circumstances because they led to the enactment of the Social Security Act of 1935, and focused worker attention, as never before, on the need for old age security measures and the generally unsatisfactory dimensions of individual savings. Moreover, the Social Security law was enacted to provide only a "floor of protection"; it was deliberately designed to leave room for private enterprise to foster initiative that would supple- ment Social Security. With the advent of World War II, the opening that had been left by the Social Security Act was widened dramatically so as to give the incentives for private pension plan expansion a significant push. The reasons for this upsurge were twofold: First, the ordi- nary and excess profits tax rates imposed on corporations during the war (and the Korean conflict as well) were extremely high. Cor- porate contributions to tax qualified pension plans were deductible and the investment income earned by pension trusts was tax exempt (also, high surtax rates on individuals made pensions at- tractive as a method of employee compensation since the tax was deferred until the income was received, usually at retirement when the marginal tax rate would be lower); second, the wage-price con- trols imposed as a wartime measure provided a specific exception for fringe benefits, including pensions, and thus permitted and en- couraged both management and labor to rechannel pressures for higher wage rates into fringe benefits. At the same time, the Revenue Act of 1942 considerably tight- ened up the tax qualification requirements for pension plans so that it no longer was acceptable to establish qualified plans only for the benefit of those in the higher ranks, and plans established during this period showed "a marked tendency to safeguard the employees' rights to benefits." 7 Despite these improvements, how- ever, by 1945 it was still doubtful that there had been any net gain in coverage over the pre-Depression years.8 For the most part, the exception created under the wage stabilization program had result- ed only in influencing the adoption of short-run fringe benefits like paid vacations, shift differentials, paid holidays, et cetera.9

- Ibid. at 930. 6 A. Epstein, Insecurity, a Challenge to America (New York: Smith & Haas, 1933), at 147-148. 7 M. Latimer and K. Tufel, Trends in Industrial Pensions (as amended) (New York: Industrial Relations Counselors, Inc., 1940) at 43. 8 S. Slichter et al., The Impact of Collective Bargaining on Management, (Washington: Brook- ings, 1960) at 373. 9 See J. Dunlop, "Appraisal of Wage Stabilization Policies" in Problems and Policies of Dis- pute Settlement and Wage Stabilization During World War II, U.S. Department of Labor, Bureau of Labor Statistics, Bulletin 1009 (Washington: GPO, 1950). The breakthrough that inspired the explosive post-World War II growth in private plans came about through new and revived trade union interest in negotiated pension programs. The Depression had caused unions to turn away from the retirement programs they had sponsored earlier and to seek resolution of the retirement income needs of their members within the framework of Social Se- curity. The stirrings of interest in negotiated plans had to over- come formidable obstacles, principally the long-held trade union belief that employer pension plans were feudalistic in character and bound the worker to the company like the serf to the manor,10 but these reservations subsided in the face of powerful develop- ments that led inexorably to the spread of negotiated plans. The first of these developments was the so-called Krug-Lewis agreement establishing the UMWA welfare and retirement fund in 1946. A proposal for such a fund was rejected by the mine opera- tors in 1945 and again in 1946, leading to a strike, a seizure of the mines by Secretary of Interior Krug and the subsequent agreement with the union. But John L. Lewis, the mineworkers' memorable leader, insisted that the program be administered exclusively by the union and the Government balked, fearing an adverse public reaction. Further strikes and litigation followed before joint labor- management administration of the fund was finally settled. Section 302 of the Taft-Hartley Act," which requires joint labor-manage- ment representation in the administration of union funds to which employers contribute, arose as a response to these circumstances. The miners' drive for an employer-financed welfare and retire- ment fund had a profound influence on other labor unions, shifting some of their attention from direct wage increases to pensions and other fringe benefits. Another important factor was the Inland Steel decision of the National Labor Relations Board (NLRB) in 1948, which held that pensions were a mandatory subject for collec- tive bargaining. Inland Steel had unilaterally established a con- tributory pension plan with a compulsory retirement feature in 1936. The company refused to bargain with the United Steelwork- ers of America on a compulsory retirement grievance on the ground that as part of the pension plan the issue was not within the collective bargaining obligation, and the union filed an unfair labor practice charge with the NLRB, alleging an illegal refusal to bargain. The NLRB upheld the union, ruling that "[w]ages must be construed to include emoluments of value, like pension and insur- ance benefits which may accrue to employees out of their employ- ment relationship." 12 The Seventh Circuit Court of Appeals upheld the NLRB but preferred to base its affirmance on the phrase "other conditions of employment" in section 8(a) of the Taft- Hartley Act.' 3 and the Supreme Court subsequently refused to grant certiorari.' 4 With the Inland Steel decision behind them, the industrial unions, having become increasingly dissatisfied with the adequacy

10See, e.g., American Federation of Labor, "Retirement Plans In Collective Bargaining," Re- search Report, October 1949, p. 3 ff. 1 29 U.S.C. § 186. 1 2Inland Steel Co. v. United Steelworkers of America, CIO, 77 NLRB 4 (1948). 13 Inland Steel Co. v. NLRB, 170 F.2d 251 (7th Cir. 1949). 4*366 U.S. 960 (1949). of Social Security benefits,' 5 began to press for negotiated pensions for their members. After prolonged and bitter strikes in the late 1940's in the steel and auto industries, the steel and auto unions broke through to win this objective. Almost concurrently, the type of pooled multiemployer fund cre- ated by John L. Lewis for the mineworkers, which actually had been pioneered by the garment trade unions back in the 1920's as a form of private unemployment insurance,16 was perceived by the craft unions as being more suited to their needs than the single- employer model fashioned in the steel and auto sectors of the econ- omy. Industries such as construction, food, apparel, mining, motor and water transportation were characterized "by seasonal and ir- regular employment, small establishments and such frequent job changes that few workers remain with a single employer long enough to qualify for pensions." 17 The pooled fund concept was the only way to cope with the great perils to which individual em- ployer plans in these industries would be exposed. Although craft unions still were giving their priority to obtaining higher wage set- tlements, by the mid-1950's, multiemployer plans had begun to spurt and in 1955, Hoffa negotiated his first pension plan for the Teamsters. The conditions giving rise to the increasingly spectacular growth of private plans were now in place. By the time the Welfare and Pension Plans Disclosure Act was enacted in 1958, private pension plans covered 17 million employees and the plans contained $27.4 billion in assets."' By 1970, the plans were estimated to cover 29.7 million employees and to hold more than $137 billion in assets.' 9 These figures may not seem that consequential today when we learn that private pension plans are now estimated to cover 35 mil- lion workers, with assets of $900 billion, and expected to grow to $3 trillion by 1995,20 yet to those who remember that the plans only covered 4 million workers and that assets of these plans had only amounted to $2.4 billion in 1940,21 the sheer size of the coverage and reserves in the late 1950's seemed staggering and an enormous plus for the Nation's economy. But there was an underside to this period of rapid growth and soon problems involving the mishan- dling of some of these huge funds surfaced. These abuses led to the first important congressional effort to regulate employee benefit funds outside the confines of the Internal Revenue Code.

15 S. Barkin, "What Shall We Have: Retirement Benefit or Superannuation Plans?" Proceed- ings, Industrial Relations Research Association, 1949 at 40; see also, S. Doc. 208, 80th Cong., 2d Sess. 2 (1949). 16 J. Seidman, The Needle Trades, 267-269 (New York: Farrar & Rinehart, " BLS, Multiemployer 1942). Pension Plans Under Collective Bargaining, spring 1960, Bulletin No. 1326, June 1962, at 1. Is U.S. Department of Labor, Legislative History of the Welfare and Pension Plans Disclosure Act of 1958, as amended by Public Law 87-420 of 1962, at 96. 19 Bankers Trust Co., New York, The Private Pension Controversy, at 9 (1973). 2o Remarks of Francis X. Lilly, Solicitor of Labor, U.S. Department of Labor, at the National Pension Forum of the ERISA Advisory Council, Washington, DC, June 14, 1984. 21 Institute of Life Insurance, Private and Public Pension Plans in the United States (March 1967). C. PROLOG To ERISA-THE WELFARE AND PENSION PLANS DiscLO- SURE ACT AND THE PRESIDENT'S CABINET COMMITTEE REPORT OF 1965 Congressional interest in the problems of improper practices in connection with employee benefit plans goes back to the early 1950's. At that time investigations were made by subcommittees of the House Committee on Education and Labor and the House Com- mittee on Government Operations. 2 2 Later, a special committee was established within the Senate Committee on Labor and Public Welfare under the chairmanship of Senator Irving Ives of New York and subsequently Senator Paul Douglas of Illinois. The Senate committee conducted wide-ranging studies of union and jointly-administered welfare and pension plans. Both House and Senate committees concentrated on abuses in the administration of these funds although in the Senate some concern was also ex- pressed with respect to the adequacy of funding for pension plans. The Senate committee investigations found that the extremely rapid growth of private pension plans had led to all manner of abuses, ranging from ineptness and lack of know-how to outright looting of benefit funds and corrupt administration. In addition to embezzlements, kickbacks, unjustifiably high administrative costs, and excessive investment of funds in employer securities, serious examples of improper insurance practices were also found, includ- ing exorbitantly high commission and administrative charges, ficti- tious fees, retention by some insurance carriers of an unduly large share of the premiums, unequal treatment of policyholders, switch- ing carriers to obtain high first-year commissions and collusion be- tween insurance representatives, union officials and manage- ment. 23 These abuses led to the Douglas-Ives bill 24 which became the basis for the Welfare and Pension Plans Disclosure Act (the WPPDA) enacted in 1958.25 The theory of the law was that full disclosure to participants and beneficiaries of the provisions of their plan and its financial operations would deter abuse ("sunlight being the best disinfectant" 26) and would enable them to police the plans themselves without requiring greater Government regu- lations or interference. In fact, the WPPDA was a greatly diluted version of the bill that had passed the Senate and deprived the Secretary of Labor of in- vestigatory and enforcement powers, leading President Eisenhower it created a precedent of to remark that it was signed only because 27 Federal responsibility and that it would have to be improved.

22 See: Hearings on investigation of racketeering before Special Subcommittees of the House Committee on Government Operations and Committee on Education and Labor, 83d Cong., 1st Sess., June 8, 11, 12 and 13, 1953; hearings on strikes and racketeering in the Kansas City area before Special Subcommittee of the Committee on Education and Labor and the Committee on Government Operations of the House, 83d Cong., 1st Sess.; hearings on investigation of welfare funds and racketeering before Special Subcommittee of the Committee on Education and Labor, of the House, 83d Cong.; hearings held at Detroit, Nov. 23, 24, 25 and 27, 1953. 23 U.S. Department of Labor, Legislative History of the Welfare and Pension Plans Disclosure Act of 1958, etc., op. cit., at 75. 24 S. 2888, 85th Cong., 2d Sess. (1958). 25 Public Law 85-836, 85th Cong., 2d Sess. (1958). The WPPDA was subsequently repealed and replaced by ERISA. See ERISA § 111, 29 U.S.C. § 1031. 26 U.S. Department of Labor, Legislative History of the Welfare and Pension Plans Disclosure Act of 1958, etc., op. cit., at 122. 27 Ibid. at 90, 99, 102. The reasons for its weaknesses stemmed from a fundamental labor- management split on the proposed scope of the law, and gave rise to a congressional decision that had important consequences for the future direction of pension legislation. Most of the abuses that had been uncovered related to multiem- ployer plans and business groups contended that these were the plans peculiarly susceptible to abuse because they had a "fixed cost contribution by the employer to a fund from which benefits are paid" as distinguished from those plans where the employer agrees to provide a "level of benefits." Efforts to limit the proposed disclo- sure law to what was tantamount to the jointly-administered Taft- Hartley funds were unsuccessful in both the Senate and the House but there was a stiff price to pay. In order to make all types of em- ployee benefit funds subject to the new disclosure requirements, proponents of the measure had to agree that the Secretary of Labor would serve as nothing more than as a warehouseman for the reports that would be filed. Nonetheless, the precedent of broad regulatory coverage for all types of plans was significant as was the selection of the Depart- ment of Labor as the administering agency. The latter was intend- ed to validate the labor interest in benefit fund regulation al- though at the time the competition came not from the IRS but rather the SEC, where the Douglas-Ives bill had originally placed administration of the disclosure provisions. 28 No sooner was the ink dry on the newly signed law, then the Department of Labor began to lobby for strengthening amendments, and with the arrival of the Kennedy administration it did not have to wait too long, since President Kennedy, as a prior member of the Senate Labor and Public Welfare Committee, had displayed an intense interest in welfare and pension plan abuse and had played an important role in the development of both the WPPDA and the union reform law known as the Labor-Management Reporting and Disclosure Act of 1959 [LMRDA],29 which also touched on many of these prob- lems. Amendments to the WPPDA were enacted in 1962 to restore most of the enforcement provisions that had been eliminated by the House in 1958. These amendments provided the Department of Labor limited investigatory authority and power to issue regula- tions. The 1962 amendments also required surety bonds of those handling plan funds and amended the title 18 criminal code provi- sions to make theft, embezzlement and kickbacks involving employ- ee benefit funds Federal crimes.3o By the time the 1962 amendments had passed, however, other de- velopments were in motion which in only a few more years were destined to cause a complete rethinking of the full disclosure con- cept as the principal means of directly regulating employee benefit plans. In 1961, a private organization known as the Commission on Money and Credit examined private pension plans and made a series of recommendations concerning their regulation, one of which called for greater supervision over the investment of pension

28 Ibid. at 77. 29 Public Law 86-257, 86th Cong., 1st Sess. (1959), 29 U.S.C. § 401 et seq. so See 18 U.S.C. §§ 664 and 1954. fund assets in the sponsoring company. 3 President Kennedy, in his 1962 Economic Report recognized these recommendations, stat- ing he had supported amendments to the WPPDA as a Senator, but that "there is also a need for a review of rules governing the investment policies of these funds and the effects on equity and ef- ficiency of the tax privileges accorded them." 32 On March 28, 1962 President Kennedy established a cabinet-level committee known as the. Committee on Corporate Pension Funds and other Private Re- tirement and Welfare Programs. It was given the responsibility of conducting: * * * a review of the implications of the growing retire- ment and welfare funds for the financial structure of the economy, as well as a review of the role and character of the private pension and other retirement systems in the economic security system of the Nation, and * * * [to con- sider] * * * how they may contribute more effectively to efficient manpower utilization and mobility. Three events were then taking place which had a powerful influ- ence on the decision to give the President's committee such a broad mandate of review and on the subsequent recommendations of the committee. The first of these was the fact that the passage of the WPPDA in 1958 had unleashed a nonstop torrent of mail from em- ployees all over the country complaining over their failure to qual- ify for private pension benefits and mistakenly assuming that the WPPDA provided some remedy in this respect. This was a phe- nomenon which did not cease until ERISA was enacted and was the basis for the enduring grassroots constituency in support of broad pension reforms. The second was the Studebaker case. In 1962, the UAW, realizing that the financial condition of the automobile producer was very shaky, had sought unsuccessfully in negotiations with the company to establish the pension plan's first claim on assets of the company in the event of a shutdown. When auto production at the South Bend, IN, facility closed in 1963, the pension plan was terminated and approximately 4,400 workers with vested pension rights under the plan-which incidentally was funded on an ongoing basis as well as any plan under ERISA's funding standards-lost some or all of their vested pensions. The incident burst like a bombshell on the private pension scene and contributed greatly to the sense that serious problems affected the plans that needed addressing. Finally, in 1964, a book by Prof. Merton Bernstein, "The Future of Private Pensions" 3 appeared, which laid out in a magisterial fashion, what Bernstein viewed as the principal deficiencies in pri- vate pension plans. Based on exhaustive case studies, he argued eloquently and persuasively that private plans did not assure equity or adequate legal protection to workers, that the plans re- stricted labor mobility, that the adequacy of funding for many

3e Money and Credit, the report of the Commission on Money and Credit (Englewood Cliffs, NJ: Prentice Hall, Inc., 1961). 32 See D. McGill, Fundamentals of Private Pensions, Third Edition (Homewood, IL: Richard D. Irwin, Inc., published for the Pension Research Council, Wharton School, University of Pennsyl- vania, 1975) at 33. 3 3M. Bernstein, The Future of Private Pensions (New York: Free Press, 1964). plans was seriously in doubt, and that termination of plans owing to mergers, corporate dissolution, financial difficulties or sales con- stituted a substantial threat to the retirement security of indeter- minate numbers of employees. Bernstein came to the conclusion that the problems of private plans were so inherently difficult to resolve that only an expanded Social Security system could provide the basic retirement security that an increasingly aging population would require. Nonetheless, making the plans more equitable and providing greater safeguards to protect the reasonable expectations of employees would still be a major step forward. Although many critics thought that Bernstein had overstated the case against private pensions, the book had an undeniable influ- ence on all who thought, wrote or spoke on the issues of pension reform. Even those who did not agree with some or most of Bern- stein's conclusions, recognized that he had provided the most com- prehensive and intellectually rigorous critique of private pensions to date, and that, henceforth, no meaningful debate on the subject could ignore his contribution. When finally issued in 1965, the Cabinet committee's report reaf- firmed the value of, and the strong public interest in, private pen- sion programs, and the importance of continuing private pensions along with Social Security and individual savings as the three-cor- nered basis for retirement income security. The report concluded that incentives should continue to be given to promote the growth of private plans but that in order to provide a more reliable foun- dation for their future development it was necessary to improve the basic soundness and equitable character of the plans. To accomplish the latter, the report made as its principal recom- mendations: (1) The imposition of mandatory minimum vesting standards, suggesting that employees be vested 50 percent after 15 years of service and 100 percent after 20 years of service. (2) The imposition of mandatory minimum funding standards, suggesting that all accrued benefit liabilities (that is, both past and current service liabilities) be amortized over a 30-year period, that, in the case of multiemployer plans which negotiated fixed contribu- tion rates, the benefits provided be limited to what the contribu- tions could actuarially embrace, that the adequacy of the funding process be certified to, at least every 3 years, by a qualified profes- sional actuary, and that the IRS construct guidelines governing the appropriateness of actuarial assumptions based on advice from an advisory board of actuaries and others. (3) That a voluntary portability system (referring to the transfer- ability of pension credits among plans) and a system of plan termi- nation insurance (called "reinsurance"), to protect against loss of vested pension benefits when a plan was terminated prior to full funding (as in the Studebaker case) be studied further to determine their feasibility. (4) That in the fiduciary area, no steps be taken to impose Fed- eral statutory standards until the effectiveness of the disclosure provisions were further tested, but that pension funds should be subject to something like a 10-percent limitation on the amount of plan funds that could be invested in employer securities; and

37-569 0 - 84 - 2 (5) Amendments to the tax code should be made to cure serious inequities in the tax treatment of pension benefits, the most impor- tant of which was the recommendation to place a dollar limitation on contributions to a plan for any employee or an equivalent limi- tation on benefits in order for a plan to maintain tax qualified status.3 4 The President's committee report excited wide discussion and controversy in the pension community. A number of noted and re- spected business and labor leaders who had served on the Presi- dent's Advisory Committee on Labor-Management Policy, to which the report had been referred, dissented vigorously from one or more of the report's major recommendations, believing that these recommendations had gone too far. Nonetheless, the report was recognized as constituting something of a watershed in developing public policy on private pensions and an interagency staff task force was set up in 1966, composed mainly of former staff from the various agencies that had functioned in connection with the Presi- dent's committee report, to draft legislation that would implement the report's recommendations. As time went on, however, it became apparent that due to the misgivings of many in business and labor circles, the political impact of the President's committee report was far less than had been anticipated, and while the inter- agency staff task force carried on its deliberations to prepare for drafting of legislation, the initiative was shifting from the execu- tive branch to the Congress. D. PERIOD OF GESTATION-CONGRESSIONAL INVESTIGATION, STUDIES, AND BILLS, 1965-72 Except for restrictions on plan investments in a plan sponsor's securities, the President's committee report had virtually dismissed out-of-hand the need for Federal fiduciary standards for private plans. In the same year that the report was released, however, the Senate Permanent Subcommittee on Investigations, a unit of the Senate Government Operations Committee, led by Chairman John L. McClellan, conducted an investigation of the Allied Trade Coun- cil and Local 815 of the International Brotherhood of Teamsters, two small New Jersey unions. This investigation demonstrated that the disclosure provisions of the WPPDA were not sufficient to deter serious abuse in the handling of employee benefit funds and that the President's committee had erred in believing that it might. In less time than it took for the President's committee report to be drafted, Federal fiduciary standards leaped to the top of the emerging pension reform agenda. What Senator McClellan's committee discovered was that one George Barasch, the founder of the two New Jersey unions in- volved, had managed to manipulate and divert the funds of the em- ployee benefit plans connected with the unions in such a way as to make himself a prospective multimillionaire. Among other things, Barasch, who was a trustee of the plans, had set up a commercial benefit consulting organization which virtually ran the plans and

3 See President's Committee on Corporate Pension Funds and Other Private Retirement and Welfare Programs, Public Policy and Private Pension Programs, A Report to the President on Private Employee Retirement Plans (January 1965) at ix-xvi ff. collected huge consulting fees for the benefit of Barasch. At the time of the McClellan investigation, Barasch was in the process of liquidating the two benefit funds and transferring $4 million of their assets to two so-called charitable corporations, established in Liberia and Puerto Rico, of which Barasch was the organizer and principal shareholder.3 5 McClellan's committee was told by executive branch representa- tives that all of the foregoing could not be prevented under existing laws. This testimony so dismayed Senator Jacob K. Javits, a member of the Government Operations Committee, that within less than 2 weeks after the committee had concluded the Barasch inves- tigation, he introduced the first bill to impose fiduciary standards on employee benefit funds. 3 6 The bill proposed to amend section 302, Taft-Hartley and contained detailed requirements as to how these trust funds should be administered. Both plan participants and the Secretary of Labor were authorized to sue in Federal court to recover losses or enjoin violations. Senator McClellan called upon the Johnson administration for technical assistance in fashioning a bill of his own and was re- ferred to the Department of Labor. The Department, in testifying on its impotence in the Barasch case, had also noted examples of similar abuse in non-Taft-Hartley employer-administered plans. Be- lieving that the Javits version might revive the labor-management split that had marred the original WPPDA, and that the problems of fiduciary abuse were not confined to the Taft-Hartley funds, the Department's technicians persuaded McClellan to introduce a bill in October 1965,37 that amended the WPPDA to set fiduciary standards for trustees of all types of plans. This bill became the forerunner of the fiduciary provisions that ultimately were enacted in ERISA, even though it did not represent official policy of the Johnson administration at the time of its introduction. Throughout 1966, the administration's interagency task force held meetings with various interest groups and drafted proposals for legislation, including a fiduciary bill. In the meantime, the UAW, concerned that the President's committee report had not made a clear cut recommendation in favor of a termination insur- ance program ("reinsurance") to handle Studebaker problems, pre- vailed on Senator Vance Hartke of Indiana to introduce a reinsur- ance bill 38 and the Senate Finance Committee held a 1-day hear- ing on the matter in 1966. This is where things stood until 1967 when on February 20 the administration unveiled its Welfare and Pension Plan Protection Act, a bill providing fiduciary and added disclosure requirements and modeled after the McClellan bill, but with important refine- ments added by the interagency task force. In introducing the bill, Senator Yarborough, then chairman of the Labor Subcommittee, stated: In order to assure ordinary care and prudence in han- dling the welfare and pension funds, persons managing the

5 See S. Rept. No. 1348, 89th Cong., 2d Sess. (1966). 36 S. 2352, 89th Cong., 2d Sess. (1965). 37 S. 2627, 89th Cong., 2d Sess. (1965). 38 S. 1575, 89th Cong., 2d Sess. (1965). funds must be responsible as fiduciaries to the funds and the participants and their beneficiaries. While the tax- exempt status of a fund may presently be lost by the trust- ees making prohibited transactions, no penalty either criminal or civil is imposed on the trustees. The extremely rapid growth of welfare and pension plans, coupled with uncertainty about the rights of the em- ployees participating in them, have afforded opportunities for abuse. Legal protection against such abuse has been de- pendent on State laws. However, under many of these laws the extent to which persons handling such funds are bound by the responsibilities and standards of "trustees" is uncertain. * * * * * * The need for a bill on this subject was made clear by the hearings conducted by the McClellan committee in the summer of 1965 and by the report of the President's Committee on Corporate Pension Funds and Other Private Retirement and Welfare Funds. While other problems in connection with such funds have not yet been solved, the urgent need for protection from mismanagement cannot be set aside.39 The statement not only summed up the rationale for fiduciary standards, but also disclosed the administration's legislative strate- gy. Fiduciary provisions had priority and appeared to command a consensus. It would, therefore, be assigned a faster legislative track and not be made to wait upon the development of further, probably more controversial proposals, in such areas as vesting and funding, that were still under consideration by the interagency task force. Senator Javits, however, had other ideas. Just a little over 1 week later, on February 28, 1967, he introduced the first broad- scaled pension reform bill, S. 1103. In addition to a section on fidu- ciary responsibility, the bill contained provisions for vesting and funding standards, a program of plan termination insurance and a voluntary central portability fund. Thus, the bill dealt with virtual- ly all the areas of concern reported on by the President's commit- tee. Moreover, the bill had been drafted in a sophisticated, knowl- edgeable manner by an experienced labor lawyer, Frank Cum- mings, who was then minority counsel for the Senate Labor and Public Welfare Committee. While the interagency task force was studying how to draft these sort of standards, Cummings had shown how it could be done. In short, with one bold stroke, Javits had taken over center stage and established the prototype pension reform bill which would end up dominating both the tone and the content of the legislative struggle that loomed ahead. In explanation of the need for such sweeping regulation, Javits said:

39 113 Cong. Rec. 3924 (Feb. 20, 1967). Although not mentioned by Senator Yarborough, an important factor bearing on the justification for Federal fiduciary standards, was the conviction of Teamster's President James Hoffa on mail fraud charges even though the substantive ele- ments of the fraud all involved pension fund transactions. The fact that Hoffa could only be successfully prosecuted under mail fraud provisions did not go unnoticed. See U.S. v. Hoffa, 367 F.2d 698 (7th Cir. 1966), vacated and remanded to District Court, 394 U.S. 310 (1967). I believe that all of these problems are so interrelated that they cannot be solved without a comprehensive legis- lative program dealing not only with malfeasance of ad- ministrators, and not only with the consequences of plant shutdowns and plant terminations, but also with the broad spectrum of questions such as adequacy of funding, reason- able minimum standards of vesting, transferability of cred- its under some circumstances, and, in short, the establish- ment of certain general minimum standards to which all private pension plans must conform. That is by no means to say that we should create a legis- lative straightjacket which would destroy the flexibility and inventiveness which have been one of the foundations of the enormous growth of pension plans in recent years. But I think there ought to be some minimum standards in this field. And, in my judgment, those minimum standards will no more force all pension plans to be the same than the minimum wage law forces all employees' wages to be the same. The minimum is merely the basic level which decency and order require. 40 The bill also established an independent SEC-type commission to administer the new regulatory provisions and Javits indicated why he thought that was necessary: The concept of a single authority to regulate the cre- ation and operation of pension plans is one which should be beneficial to the labor organizations, employers, and participants as well as those who sell and service pension plans, such as the banks, insurance companies, and pen- sion consultants. Presently, various aspects of some pen- sion plans are unregulated while other aspects are regulat- ed by the State agencies, by the Treasury Department, Labor Department, Securities and Exchange Commission, among others. The need for coordination of these efforts together with any new regulation is obvious. * * * * * * The point to be made here is that a great deal of thought will be required to develop a rational and coordi- nated system for the regulation of pension and other em- ployee benefit systems without adversely affecting the tra- ditional role of existing agencies now concerned with some aspect of these plans. But the goal is an important one, and worthy of the effort that will be required. For, after much thought on this subject, I am convinced that a single agency is required. It will be a very difficult task to regu- late the operation of employee benefit plans sufficiently to assure the legitimate expectations of employee partici- pants while at the same time avoiding undue or unneces- sary interference with the operation of these plans. Over regulation or unnecessary regulation would be worse than none, for it would deter the installation and improvement of these much needed programs. It is a tortuous course to be steered between the problems of frustrated expectations

40 113 Cong. Rec. 4650-4653 (Feb. 28, 1967). for pension plan members growing out of no regulation and the equally damaging frustrations growing out of an irrational regulatory scheme which deters the employer from instituting a pension plan for his employees. It seems clear to me that this course could best be steered by an agency which has the general responsibility for encourag- ing the growth of the private pension system including the implementation of needed rules to govern pension partici- pants.4 1 The introduction of the new Javits bill startled the administra- tion and galvanized the interagency task force into speeding along the drafting process. The issue of administration of new standards was finally resolved when Assistant Secretary of Treasury Stanley Surrey reported that Chairman Wilbur Mills of the House Ways and Means Committee had refused to handle the proposals because of business opposition. It was decided that the proposals would have to function as a labor bill rather than a tax bill, with admin- istration given over to the Department of Labor. In substance, the task force decided that the recommendations of the President's committee report were too conservative, not up-to- par with the more progressive private plans, and, if translated without change into a legislative proposal, would not compete suc- cessfully with the Javits initiative. Accordingly, faster vesting re- quirements (10 years after reaching age 25), faster funding (funding of vested liabilities within 25 years) and establishment of a corpora- tion within the Department of Labor to administer a termination insurance program were all adopted. The only matter left to fur- ther study was the portability issue. A bill incorporating these decisions was not introduced, however, until May 1968.42 In the interval between introduction of the Javits bill and May 1968, a serious deterioration had occurred in the relationship between Secretary of Labor Willard W. Wirtz, who had now assumed primary responsibility for pension legislation within the administration, and President Johnson. In addition, the latter had been advised that a Presidential endorsement of the interagency task force bill would cost him business support if he chose to run for reelection (which, as it turned out, he didn't). The result of all this was that when the interagency task force bill was introduced in 1968, it was denied the administration's formal spon- sorship and was widely believed to reflect an act of rebellion by Secretary Wirtz, who characterized the bill as a Department of Labor initiative. Although Senate hearings were held on the bill shortly after it was introduced,4 3 without administration support it never got off the ground, and when the Nixon administration came into office, and asserted that the pension proposals, except for fidu- ciary and disclosure amendments, required much further study, and hinted that it was unlikely that vesting and funding reforms

*' Ibid. 42S. 3421, 90th Cong., 2d Sess. (1968). For introductory remarks and text, see 114 Cong. Rec. 11544-11551 (May 2, 1968). 4 3 Hearings on Pension and Welfare Plans before the Subcommittee on Labor of the Senate Committee on Labor and Public Welfare, 90th Cong., 1st Sess. 1968). would be considered favorably, the movement for broad pension re- forms came to a complete halt. E. THE DRIVE FOR ENACTMENT Once again, Senator Javits proved to be the catalyst that got things off of dead-center. In December 1969, Joseph ("Jock") Ya- blonski Sr., the unsuccessful challenger in a bitter election contest for the presidency of the UMWA (which was under protest to the Department of Labor), had been mysteriously murdered, setting off a nationwide clamor for an investigation into the affairs of the union. The Senate Labor Subcommittee was assigned the task but since one of the charges that had been made by the Yablonski fac- tion against the incumbent president of the Mineworkers, W.A. ("Tony") Boyle, was that he had misused the UMWA health and retirement funds in connection with the disputed election, Javits seized the opportunity to insist that the resolution establishing the investigative authority of the Labor Subcommittee, also require a ''general study of pension and welfare funds with special emphasis on the need for protection of employees covered." 44 The chairman of the Labor Subcommittee, Harrison A. Williams, Jr., of New Jersey agreed, and in March 1970, Senate Resolution 360 was en- acted, launching the study.45 At the time, the pension study appeared to be a very small tail on a very large UMWA dog and it did not receive much attention. But as things turned out, it was a pivotal maneuver-one that was to awaken the conscience of the American public to the inequities in private pensions. At the same time, the study defined the pen- sion reform agenda in a manner that freed it from the arid dis- courses of blue ribbon committees, academicians and lawyers, all of which had been useful, but which had failed to attract popular in- terest. In due course, the study made private pension deficiencies into a media event. The key happening was the release in March of 1971 of a contro- versial segment of the study dealing with historical forfeiture rates in a small sample of specific pension plans which had been sur- veyed by the staff of the Labor Subcommittee. 46 Translated into plain English, which the newspapers were quick to do, forfeiture meant loss of eligibility for anticipated retirement benefits because of an employee's failure to meet a plan's age and/or service re- quirements, usually the latter. The study found that in the sample of plans studied, which had lengthy service requirements, only 5 percent of the millions of employees covered since 1950 had ever received benefits, only 8 percent had qualified for benefits, and while most of these employees had only worked a very short period of time (less than 5 years), there were substantial numbers of work- ers who had longer periods of service and failed to qualify for bene- fits. These findings created an overnight sensation and caused a furor among pension experts which smolders to this day. The study was roundly criticized on a variety of grounds by respected authorities

4 4 See S. 4 5 Rep. No. 92-634, op. cit., at 1. 4 6 Ibid. 1d. at 119. but the findings seem to dovetail with the personal experiences of a great number of people, and certainly with the letters of complaint that had been pouring into Government offices for so long. Those experiences and letters could be summarized very simply: you could work a long time for a company and still not get a pension because the rules were so strict or out-of-date; only a lucky few made it to the end and qualified, and it did not matter whether you were blue collar or white collar, rank-and-file or executive. The same trap awaited all. What the President's committee and Professor Bernstein had known and described was now knowledge possessed by the Ameri- can public and it cut very deep. Along with fiduciary standards, vesting requirements had achieved priority status in the eyes of pension reformers. Opposition to the possibility of federally-mandated vesting stand- ards took the form of counter-studies which attempted to show that private plans had made great strides in achieving adequate vesting and that mandated vesting for those who were backward in this re- spect would be too costly. The latter was the more significant con- tention and so, eventually, the Labor Subcommittee commissioned an independent actuarial study to determine the projected cost impact of universal vesting.47 At the same time, the subcommittee began to examine the case for reforms in the areas of funding, ter- mination insurance, portability and augmented disclosure. The scope of these activities not only constituted an endorsement of Senator Javits' comprehensive approach to pension reform, but also served to deflect pressure from proponents of more limited leg- islation, such as the Nixon administration, which wanted the sub- committee to report only a fiduciary-disclosure bill, and the auto and steelworkers unions, who wanted the subcommittee to concen- trate on passing a termination insurance proposal. In fact, in De- cember 1970, Senator Javits, much to Senator Hartke's chagrin, had prevented the latter from successfully attaching a termination insurance amendment to Social Security legislation on grounds that such an amendment was premature and would not be feasible unless tied to federally imposed funding standards.48 Hartke and the unions were never quite convinced that Javits could deliver the comprehensive reforms he was seeking and were concerned that in the process the opportunity to achieve plan termination protection would be lost. Throughout 1971, the Labor Subcommittee held hearings on the principal areas of purported plan deficiencies. The initial hearings in July of that year 4 threw the spotlight on individual cases of failure to qualify for retirement benefits despite long years of serv- ice. A succession of workers in their late forties and fifties told of losing their pension rights because of layoffs, plant shutdowns, transfers and business closings. One witness, for example, had worked 32 years for one company and was laid off 3 years before he was eligible for his pension, although, subsequently, even more

4 Id. at 113. For the actuarial study itself, see S. Rep. 92-1150, 92d Cong., 2d Sess., appendix 11 (1972). 4 See 116 Cong. Rec. 40877-40889 (Dec. 10, 1970). 4 See hearings, Private Welfare and Pension Plan Study, 1971, before the Subcommittee on Labor of the Senate Committee on Labor and Public Welfare, 92d Cong., 1st Sess. (1972). atrocious examples surfaced. The hearings also dealt with some in- stances of underfunded plan terminations and reviewed with work- ers who had been involved, the dimensions of the Studebaker trage- dy. These "horror stories," as they were referred to, captured media attention and solidified public opinion in favor of vesting reforms. In October 1971, the Labor Subcommittee heard the plan sponsor's side of the story. Although some company representatives defended their failure to provide more progressive vesting on grounds that their first concern was improving retirement benefits for older em- ployees, implying that there was conflict between the two goals, the majority said it was a question of costs, even though a number of the plans involved had been overfunded. The October hearings also turned up instances of substantial un- derfunding in both single and multiemployer plans as well as misuse, manipulation and poor management of trust funds. One fi- nancially ailing company tried to borrow over a million dollars from a subsidiary's pension pool for use as operation capital. An- other had a policy of investing more than half its pension funds' assets in the company's own common stock and in the real estate of a company subsidiary. Still another company routinely dipped into its pension funds for cash to make acquisitions. Generally, the testimony presented attempted to minimize the hazards or problems presented by these practices, and for the first time, but not for the last, multiemployer plan representatives took the position that vesting requirements were not necessary in con- nection with their plans by reason of the employee's ability to transfer employment from one participatory employer to another without losing credit for previous service. What would happen if an employee left the scope of the plan prior to achieving vested status was left up in the air. Despite the efforts to show that the problems that had surfaced were rare or isolated and did not warrant legislative treatment or would, if regulated, invite excessive governmental interference, the case for sharply limited reforms was gradually unraveling. Follow- ing the subcommittee's hearings, Senator Williams agreed with Senator Javits to work out a joint pension reform bill that would incorporate the principal elements of the Javits bill. In November 1971, the White House Conference on Aging met and issued a report in December recommending that the Federal Government require early vesting and/or portability, fiduciary responsibility, funding standards and termination insurance in private plans. The recommendations seemed to constitute a mirror image of the Javits proposals. Although probably not so intended at the time, the final blow turned out to be new Nixon administration proposals deliv- ered in December 1971, virtually at the same time as the issuance of the report of the White House Conference on Aging. The new administration initiative endorsed a federally mandated vesting standard, albeit one considerably weaker than the one pro- posed by Javits,5 0 but the administration continued to stoutly oppose funding and plan termination insurance requirements 5 1 But, the big news was a major new proposal in the tax area, one that was clearly conceived as the centerpiece of the initiative and destined to become the administration's one truly popular proposal. In addition to.raising the tax deductible limits allowed to self-em- ployed persons for funding their retirement benefits under so-called Keogh or H.R. 10 plans, the administration recommended the es- tablishment of counterpart plans for employed individuals who lacked coverage under an employer or union sponsored pension program. Although the tax deductible limits for these individual retirement account plans (IRA's) were considerably lower than those provided under Keogh plans, the administration reasoned that the self-employed were entitled to a greater deduction because their plans were required to cover their employees in order to obtain tax qualification. The administration's new IRA tax proposal, which had originat- ed and been put in practice in Canada, 5 2 received considerable at- tention as a means of expanding the opportunity for private pen- sion coverage. But it also occurred to the politically astute that the administration had hit on a way to head off the stronger reforms by Javits which were likely to be embraced by Williams. 5 Admin- istration spokesmen argued that expanding coverage of private plans took precedence. After all, they pointed out, even if they complained about it, at least those workers covered by a plan- had something while those not covered had nothing. By packaging its vesting proposal in an attractive tax bill it was apparent that not only did the administration hope to bypass an unfriendly labor committee but also to shift the terms of the pension debate as well. Moreover, the IRA proposal offered an opportunity for building re- tirement income to those highly-mobile professional employees, such as aerospace engineers and technicians, who were unlikely to achieve vesting even under the liberal vesting schedule in the Javits bilL Obtaining the support of such a highly-educated, sophis- ticated and articulate employee group would be no small achieve- ment.54 Thus, by 1972, battle was joined between the administration and the Senate Labor Subcommittee on the substance and scope of the reforms that should be legislated, including the jurisdictional basis for legislating the reforms, but the administration had made an im- portant concession by offering vesting proposals. What was left was funding, termination insurance, and portability, and the prospects

5 0 The administration endorsed the "rule of 50," i.e., when an employee's age and service added up to 50 he or she was 50 percent vested, and 100 percent vested 5 years thereafter. See House Doc. No. 92-182, 92d Cong., 1st Sess. (1971). 51 See testimony of Hon. J. D. Hodgson, Secretary of Labor, hearings on S. 3598, part 1, before the Subcommittee on Labor of the Senate Committee on Labor and Public Welfare, 92d Cong., 2d Sess. at 94 (1972). 52 See testimony of Norman H. Tarver, hearings on S. 3598, op cit., part 2 at 503-4, 512. * See, e.g., E. Landau, W. Merholtz, H. Perkins, Protecting A Potential Pensioner's Pension- An Overview of Present and Proposed Law on Trustees' Fiduciary Obligations and Vesting, 40 Brooklyn Law Review 521 at 563 (1974). 54 To deal with the concerns of professional employees, such as aerospace engineers, the Senate Committee on Labor and Public Welfare subsequently adopted in 1973 provisions which ultimately became ERISA § 3032. of the administration bending on the latter two subjects did not appear very bright. In February 1972, the Senate Labor Committee published its in- terim report and in its "findings" section listed the following defi- ciencies in private pension plans: (a) Inadequate or nonexistent vesting provisions which result in the denial of retirement benefits to employees upon termination of services, voluntary or otherwise, de- spite long years of employment. (b) Inadequate accumulation of assets in funds to meet obligations and payments to workers who are entitled to benefits. (c) Forfeiture of earned retirement benefits by employees resulting from a voluntary or involuntary move from within an industry or geographical area. The failure to ac- knowledge and permit transfer of pension credits from one employer to another results in denial of all accrued bene- fits which the employee may have acquired through former employment. This restricts the mobility of the labor force. (d) Instances where employers have not achieved full funding status, but through circumstances often beyond their control, must terminate the plan without adequate resources for payment of benefits due. (e) The lack of uniform requirements of conduct by fidu- ciaries and employers in the administration and operation of their pension funds which results in abuses and un- sound practices which jeopardize the security of the assets and threaten the availability of funds for employees. (f) Employee participants not having full comprehension of their rights and obligations under their participating pension plans because of inadequate communication to them in booklets or other format of details of plans. They are not adequately informed with respect to acts or omis- sions by them which result in disqualification from or qualification for plan benefits.55 The report also made the following legislative recommendations: (1) A Federal law establishing minimum standards of vesting. (2) A Federal law establishing systematic requirements for funding of pension plans, accompanied by a program of plan termination insurance to cover unfunded vested bene- fits. (3) A uniform Federal standard of fiduciary responsibil- ity. (4) A Federal law requiring improved disclosure and communication of plan provisions to workers to be accom- plished in part by the revision of existing disclosure re- quirements to more effectively secure this objective.

55 See S. Rep. No. 92-634, op. cit., at 112. (5) The institution under Federal guidelines of a pro- gram to develop portability and reciprocity among private pension plans. (6) The centralization in one agency of all existing as well as prospective regulation of private pension plans, to the maximum extent feasible. 56 The Javits comprehensive approach was now official doctrine of the Labor Subcommittee. At this point, the chairman of the House Labor Subcommittee, Congressman John Dent of Pennsylvania and the ranking minority member, John Erlenborn of Illinois, agreed to commence a pension study of their own with a view toward formulating appropriate reform legislation. This was a signal that reforms were now regard- ed as serious and inevitable although the position of the House Labor Subcommittee on the nature of those reforms remained un- certain for many months to come. In March and April 1972, following the interim report, the Labor Subcommittee held field hearings in St. Louis, Newark, Minneapo- lis, Cleveland, and Philadelphia, exclusively on the problems of un- derfunded plan terminations and their adverse effect on employ- ees.5 7 This was done to demonstrate that while Studebaker prob- lems were relatively infrequent, when they occurred they had a devastating impact that, contrary to the administration's view, ne- cessitated a legislative response. In May of 1972, Williams and Javits, after months of staff negoti- ations, introduced a joint bill 58 which henceforth replaced the original Javits proposal. There were many modifications and im- provements in the joint bill but the basic Javits structure and ap- proach was retained with one notable exception: Javits was com- pelled to relinquish the concept of an independent SEC-style com- mission as the centralized administrator for private pension regula- tion, and accepted the Department of Labor instead. Williams had pointed out that both business and labor, for differing reasons, op- posed administration by an independent regulatory agency, and since business was bound to oppose the substantive portions of the bill anyway, the bill had no chance of succeeding without labor support, which was conditional on Department of Labor supervi- sion. The Labor Subcommittee now determined to probe the depth of support for the new Williams-Javits bill and in June held legisla- tive hearings on both that bill and the administration's proposal, even though, technically, the latter was not before the subcommit- tee. In general, industry representatives gave their support to the administration's proposals and labor to the Williams-Javits bill. Al- though a year before, industry had opposed any sort of mandated vesting on grounds it would discourage the voluntary establish- ment of private plans, it now argued that the vesting schedule of the Williams-Javits bill (which called for vesting in steps, starting

"5Ibid at 113. . See hearings, Private Welfare and Pension Plan Study, 1972, before the Subcommittee on Labor and the Senate Committee on Labor and Public Welfare, parts 1, 2, and 3 (1972). 58 See Subcommittee on Labor, Senate Committee on Labor and Public Welfare, Retirement Income Security for Employees Act of 1972, S. 3598, 92d Cong., 2d Sess. (Committee print, 1972). with 30 percent at the end of 8 years and reaching 100 percent at the end of 15 years) would excessively burden the plans whereas the administration's vesting proposals would not. As for labor, it failed to present a united front, with representa- tives of the craft unions continuing to maintain that their multi- employer plans should be exempt from vesting, causing Javits to confront them with a sheaf of letters written by former partici- pants in multiemployer plans who had complained of failure to qualify for vested benefits despite long years of service.59 At the same time, labor, spearheaded by busloads of steelworkers, over 700 of whom crowded into the hearing room, and even some man- agement representatives, complained that the Williams-Javits bill was too cautious in restricting vesting and termination insurance benefits to service earned after the date of the bill's prospective en- actment, thus failing to provide adequate protection to those who already had long years of service behind them, and urged that it be changed to cover pre-enactment service as well. In addition to divisions within the ranks of labor, there were also some nasty jolts from what had been presumed to be the natural allies of the Williams-Javits bill. Professor Bernstein, the influen- tial critic of private pension plans, testified that the Williams- Javits bill, although a positive step in the right direction, was really nothing more than a "mirage," and that if enacted it would turn out to be a cruel hoax. Bernstein advocated much faster, earli- er vesting and the imposition of stronger requirements in almost every other respect.6 0 The same was true of Ralph Nader. Earlier in the year, he had condemned private pensions as a "comprehensive fraud" and pro- posed restructuring all private pension plans into a limited series of SEC-licensed mutual funds which would invest employer and employee contributions and provide annuities at retirement.6 1 The Nader proposal was modeled after the TIAA-CREF retirement plan for college teachers, and like that plan, would have provided immediate vesting 6 2 and total portability, as employees could switch their retirement savings from one licensed fund to another. Although he did not testify at the hearings, Nader criticized the Williams-Javits bill as being nothing more than a product of that "curious pension coalition of industry and organized labor." 63 In fact, such a coalition was nowhere in sight, and in the months after the hearings the staff labored to revise the Williams-Javits bill in the hope of strengthening its basis of support. This effort was assisted materially by the release in early September of the subcommittee's independent actuarial study of the costs of mandat- ed vesting, which showed that, overall, the cost increase that would result from legislating vesting under either the William-Javits bill or the administration's bill would be quite modest, that that result would not be significantly affected in most instances by giving

59 See hearings on S. 3598, op. cit., part 3 at 1116-1143 (1972). 60 Ibid., part 1, at 296. 61 See R. Nader, A Comprehensive Fraud, 8 Trial 13 (November-December 1972). 62 In fact, the immediate vesting in the TIAA-CREF plan is somewhat misleading because participation was generally restricted until age 30. See hearings on S. 3598, op. cit., part 2 at 489. 6 3 See R. Nader, A Comprehensive Fraud, op. cit. credit for preenactment service, and that, if anything, the adminis- tration's vesting proposal, which vested older workers faster than younger ones, might be more expensive than the Williams-Javits bill. 6 4 Somewhat later, the administration's own cost study con- firmed these findings, with the result that business groups started to distance themselves from the administration's proposal, al- though it was never entirely abandoned, and began to dispose themselves more favorably toward the Williams-Javits bill. Armed with these findings and having made substantial revi- sions to the bill to deal with the problems raised at the hearings, the Williams-Javits bill was reported unanimously by the full Senate Committee on Labor and Public Welfare in mid-September 1972, the first congressional committee to report a comprehensive pension reform bill. The report accompanying the bill reviewed the inequities and abuses that had led to the bill's adoption, analyzed the inadequa- cies of existing regulation, and justified the imposition of compre- hensive new requirements on a voluntary private pension system in the following manner: The committee enthusiastically endorses the concept of a comprehensive private pension reform program. It be- lieves that expeditious enactment of S. 3598 will institute a program which will achieve a strengthening of the role played by private retirement plans within the fabric of our economic and social structure. Its most important purpose will be to assure American workers that they may look forward with anticipation to a retirement with financial security and dignity, and without fear that this period of life will be lacking in the necessities to sustain them as human beings within our society. The enactment of pro- gressive and effective pension legislation is also certain to increase stability within the framework of our Nation's economy, since the tremendous resources and assets of the private pension plan system are an integral part of our economy. It will also serve to restore credibility and faith in the private pension plans designed for American work- ing men and women, and this should serve to encourage rather than diminish efforts by management and industry to expand pension plan coverage and to improve benefits for workers. The committee believes that the legislative approach of establishing minimum standards and safeguards for pri- vate pensions is not only consistent with retention of the freedom of decisionmaking vital to pension plans, but in furtherance of the growth and development of the private pension system. At the same time, the committee recog- nizes the absolute need that safeguards for plan partici- pants be sufficiently adequate and effective to prevent the numerous inequities to workers under plans which have resulted in tragic hardship to so many. The committee has vividly demonstrated this need in public hearings.

64 S. Rep. 92-1150, appendix II, op. cit. The bill reported by the committee represents an effort to strike an appropriate balance between the interests of employers and labor organizations in maintaining flexibil- ity in the design and operation of their pension programs, and the need of the workers for a level of protection which will adequately protect their rights and just expectations. In adopting this approach, the committee believes it has designed a bill, which, like the National Labor Relations Act, the wage-hour laws and other labor standards laws, brings the workers' interests up to parity with those of em- ployers. This legislation strikes an appropriate and equita- ble balance between two opposing schools of thought- those who advocate complete and stringent control of pri- vate pensions and those who oppose any form of govern- ment supervisory or regulatory control. 65 In short, the bill, in the eyes of the committee, represented middle-of-the-road reforms, doing just what was thought germane and necessary to get the job done. In order to improve the basis of its support, the committee had adopted a large number of significant changes, among them provid- ing a form of coverage for preenactment service with respect to both vesting and termination insurance, permitting plans to opt for 100 percent vesting after 10 years of service, requiring every plan to provide a reasonable claims review procedure and providing some earlier effective dates with, however, the opportunity to obtain delays when there were economic problems. The committee had also held the line on multiemployer plans, granting no excep- tion from vesting requirements. 66 Based on the unanimous committee recommendation in favor of the bill, both Williams and Javits expected it to be taken up and passed by the full Senate before the end of the session. However, even though the bill had been deliberately drafted as a labor law and avoided amending or referring to the Internal Revenue Code, the Senate Finance Committee requested a referral of the bill on grounds that it affected tax jurisdiction. Within 10 days, thereafter, spurred on by a coalition of business groups and the administra- tion, the Finance Committee had gutted the bill of all its signifi- cant reforms, stating that it believed that "coverage, vesting, fund- ing and related provisions should continue to be dealt with by the tax committees of Congress. 67 Javits went to the floor of the Senate and delivered one of the angriest speeches of his legislative career, berating the leadership of the Finance Committee, saying it would take "a magician to demonstrate" how their action would serve the country and the American worker, and claiming that their point of view was "myopic" and based on "reactionary opposition". 6 8 The action of the Finance Committee killed chances of passing pension reforms later that session but Javits' speech ignited a fire-

so Ibid. at 13. 66 Id. 13-15, 17-26, 32. 67 Javits, Jacob K. The Autobiography of a Public Man (Boston: Houghton Miffin Co., 1981) at 383. 8 Ibid. storm of criticism against the Finance Committee, and their rejec- tion of the Williams-Javits bill became a blessing in disguise. Within weeks, letters of protest poured into Congress from all over the country and editorial writers had a field day. When Congress recessed to campaign during the Presidential election year, many Congressmen and Senators, who had never heard of the pension bill, discovered a great many disturbed and resentful constituents on their hands for whom they had no ready answers. It was apparent that the enemies of the Williams-Javits bill had made a serious miscalculation, and that the issue of private pen- sion reform had truly captured the public imagination. Now there was little room for doubt that a broad consensus had formed behind the comprehensive reforms adopted by the Committee on Labor and Public Welfare. Action to achieve those reforms could no longer be delayed or sabotaged. In December 1972, the staff of the Joint Committee on Taxation, the professional tax resource of both houses of Congress, made discreet contacts with the staff of the Committee on Labor and Public Welfare to explore a joint labor- tax bill embodying the Williams-Javits reforms and some impor- tant tax changes. The case for ERISA had been successfully made. F. EPILOG-REFLECTIONS ON THE SPIRIT OF ERISA Despite the budding rapprochement between the Senate Labor and Finance Committee, pension reform legislation was by no means out of the woods in 1973. Many hurdles remained to be over- come before ERISA became a reality. The fate of termination in- surance still hung in the balance. The resolution of the labor-tax jurisdictional issue remained the most delicate of many delicate problems. The fiduciary provisions, long taken for granted as set- tled, came under searching scrutiny and turned out to be far from settled. New provisions emerged, such as joint and survivor bene- fits, and stirred new debates. Entire sections of the Internal Reve- nue Code pertaining to pension plans were to undergo extensive re- vision and arouse even more heated controversy than some of the original Williams-Javits proposals. In brief, the picture became more, rather than less, complicated. Nevertheless, in retrospect it is clear that rejection of the Labor Committee bill by the tax committee in 1972 had beneficial effects beyond the unintended one of fortifying the consensus for compre- hensive reforms. In many respects, the Williams-Javits bill ap- proved by the Labor Committee was an unfinished and unduly cau- tious product, a result of fighting for its life at virtually every step of its development. Once the Finance Committee grasped the inevi- tability of sweeping reforms, it contributed greatly to the improve- ment of the concepts originated by the Williams-Javits bill and, with the assistance of its tax experts, was able to streamline many provisions, expand others and solve problems that the Labor Com- mittee had sidestepped or overlooked. In the end, the bill became much more complex and dense, one given over to the elaborate mechanical formulae favored by tax draftsmen and their constituencies, and one greatly encrusted with special favors for those few whose influence seemed to greatly matter. But it was also a much better bill than the one that the first pension reformers had authored. What sustained the entire enterprise, and continues to sustain it now, was the original pension reform vision Javits put forth in 1967, not an especially revolutionary vision but one that could com- mand, and still commands, a broad national consensus. This vision was summarized by Javits, 2 weeks after President Ford signed ERISA into law on Labor Day 1974: The problem, as perceived by those who were with me on this issue in the Congress, was how to maintain the vol- untary growth of private plans while at the same time making needed structural reforms in such areas as vest- ing, funding, termination, etc., so as to safeguard workers against loss of their earned or anticipated benefits-which was their principal cause of complaint and which-over the years-had led to widespread frustration and bitter- ness * * * the new law represents an overall effort to strike a balance between the clearly-demonstrated needs of workers for greater protection and the desirability of avoiding the homogenization of pension plans into a feder- ally-dictated structure that would discourage voluntary initiatives for further expansion and improvement. 69 But, as Javits pointed out, the vision was not a static one, for he concluded by saying: The need to make private pension plans more socially responsive does not stop with the enactment of the new law. The need to improve the portability provisions and the vesting in order to stimulate greater qualification by workers for pension benefits will become apparent. Also, attention will have to be devoted to providing some mecha- nism to safeguard hard-earned pension benefits against erosion from inflation. The investment policies of pension funds will have to be investigated thoroughly to determine the degree to which these huge funds can function more responsively to the imperatives of our economy. And, we also have to examine closely the effectiveness of integra- tion of private pensions with social security. All these problems are quite complex and require consideration. In short, the Congress has made an auspicious beginning with the enactment of the pension reform bill but there still remains a great deal to do if we are to promote a more satisfactory private retirement system-one that will enable every American, after his or her productive years, to look forward to a retirement with freedom from anxiety and economic want.7 0

69 Address by Senator Jacob K. Javits, Briefing Conference on Pension and Employee Bene- fits, New York State School of Industrial and Labor Relations, Cornell University and Federal Bar Association, Washington DC (Sept. 19, 1974). 70 Ibid.

37-569 0 - 84 - 3 Chapter 2 THE GOALS OF ERISA AND THE IMPACT OF ERISA ON PLAN PARTICIPANTS (Prepared by Thomas C. Woodruff,' Ph.D., Executive Director, Commission on College Retirement) A. PENSIONS BY THE 1970's Independent retirement living supported by pension benefits is a relatively new phenomenon. Only in the latter half of the 20th cen- tury has this way of life become possible for the majority of those who reach retirement age. The growth of our retirement income structure had been dramat- ic by the early 1970's. The retirement income structure in the United States was composed of a wide variety of public and private retirement income programs. The Social Security system was the broadest program, covering about 90 percent of the work force. There were several hundred thousand employee pension plans, over 6,000 State and local employee pension systems, and over 60 Federal employee pension programs. Growth in employee pension plans was significant during the 1940's and 1950's, in part due to a combination of high corporate profits, excess profits taxes, and the tax advantages associated with establishing a pension plan. However, other factors were important to the growth of employee pensions. Pressure from labor organizations whose wage demands were re- stricted by Second World War labor policies led to demands that management negotiate over pensions. As a result, pensions became a fringe benefit that both labor and management viewed as desira- ble. Private pension coverage of non-Government wage and salary nonagricultural employment grew from 14.5 percent in 1940 to 25 percent by 1950. This figure continued to increase to nearly 41 per- cent by 1960 (see table 1). By then, basic industry and unionized companies were nearly all covered by some form of pension plan. Pension growth slowed after that, reaching 45 percent by 1970 and 48.6 percent by 1975. It should be noted that these figures exclude even fulltime workers in the agricultural sector, a low pension cov- erage segment of the population. While this exclusion may have been justified in the past, when the family farm was the agricultur- al model, their exclusion today seems unwarranted.

1The views expressed in this chapter are those of the author alone and not of the Commission -on College Retirement. TABLE 1.-PRIVATE PENSION PLAN COVERAGE (ACTIVE WORKERS, BOTH FULL TIME AND PART TIME, EXCLUDING SELF-EMPLOYED) [In percent]

PresidentsCommision onPension OL/SSAcurrent population survey Age,years on job,average weeldy hours Policyrosehok ourvey 1919 Male Female emplol Male Female employees

Total...... 42 51 32 43 50 31 Underage 25...... 27 33 20 19 22 15 Age25 andover...... 47 56 36 52 60 38 ERISAstandards (morethan 1 year of serviceplus 1,000 hours)...... 58 64 48 61 67 50

SourcePresident's Commission onPension Policy, Department of Labor, Social Security Administration, 1979. While it is impossible to estimate how much of the growth in pri- vate pension coverage is due to tax treatment alone and how much to other factors, its growth since 1940 has been impressive. Howev- er, the rate of pension growth has slowed significantly. As coverage passed 40 percent, the rate of plan formation in proportion to the labor force leveled off. Much of the gain in absolute numbers of people covered can be attributed to the overall growth of the labor force. Prior to ERISA, four Federal regulatory structures were in place, as well as a number of State laws affecting the administration of pension plans. The major Federal regulatory structure rested with the Internal Revenue Service. In order to qualify for favorable tax treatment, plans had to operate for the exclusive benefit of plan participants and beneficiaries and had to meet other IRS standards. A substan- tial body of experience developed with the issue of pension plan dis- crimination. The IRS standards were, however, generally limited to a traditional concern of that agency: ensuring that taxpayers do not take excessive exclusions from their taxable income. Adequate pension funding was not a major concern. Sections of the Taft-Hartley Act, passed in 1947, were intended to address the problem of union domination of certain negotiated pen- sion plans. The act established that where unions were involved in the administration of pension plans, there had to be equal employ- er representation. By the 1970's, however, enough cases of abuse were evident that many in Congress felt that the structural re- quirements of Taft-Hartley were not adequate. Another Federal legal enforcement aim was criminal law. Em- bezzlement, the filing of false reports, and bribery were felonies. However, the Department of Justice found that it was difficult to get convictions and that when it did, they were ineffective in cor- recting abuses. The States also were not effective in regulating pension plans. Their sometimes conflicting laws made multi-State employers and plans more and more difficult to manage and operate. Conflicting State laws made it difficult for multi-State unions and employers to design and manage their pension plans. In 1958, Congress passed the first reporting and disclosure law, the Welfare and Pension Plans Disclosure Act [WPPDA]. By 1970, this statute was perceived by many to have been the greatest disap- pointment of all. After 10 years, WPPDA had shown that mere re- porting and disclosure was insufficient to permit pension plan par- ticipants to reform their pension plans on their own. The collapse of the Studebaker Corp. in 1964 helped bring togeth- er the various forces seeking reform in private pension plans. When the company declared bankruptcy, even fully vested, long service employees lost substantial pension benefits. President John Kennedy had already established a Cabinet com- mittee to study the problems of private plans. When the committee issued its report in 1965, it called for a number of reforms includ- ing: fiduciary standards, funding standards, plan termination in- surance, improvements in benefit eligibility, and a national porta- bility clearinghouse. The administration's first pension reform proposals were intro- duced in 1968. Dozens of pension reform bills were introduced in the 89th, 90th and 91st Congresses. The pension reform package that emerged in 1974 contained all of the elements of the Presi- dent's Cabinet committee report except for the portability clearing- house. B. THE GOALS OF ERISA ERISA represented a departure from the past in that it sought a coordinated set of approaches to achieve increased benefit security for a greater portion of the private sector labor force. Unlike past, piecemeal approaches, ERISA addressed the problems facing pri- vate pension plans with the following weapons: fiduciary standards, funding standards, reporting and disclosure rules, and plan termi- nation insurance, in addition to new tax-qualification rules.

1. FIDUCIARY STANDARDS ERISA clearly establishes plan trustees as the individuals who ensure that the pension plans are well-administered and will meet their pension promises. They are to act "solely in the interest" of the pension trusts' beneficiaries. ERISA modified the "prudent man rule" of common law govern- ing the behavior of fiduciaries (plan trustees and others) and re- quired that the trustees manage the pension plan for the "exclu- sive benefit" of its participants and beneficiaries. The Department of Labor later issued "prudence" regulations further defining the meaning of ERISA's modification of the "prudent man rule." Those regulations embraced modern portfolio theory for the management of pension fund assets.

2. FUNDING STANDARDS Prior to the passage of ERISA, many plans operated on essential- ly a "pay-as-you-go" basis or by contributing interest-only pay- ments on their unfunded benefit liabilities to their pension trusts. Simply put, ERISA required that current obligations should be funded in the year in which they occur, and "past-service" obliga- tions should be amortized over a 30-year period (longer periods were permitted for certain collectively-bargained plans). The 30- year amortization period was a compromise reached between those who wanted more rapid funding and those who were concerned that many employers could not afford to increase their contribu- tion rates to meet a more rapid schedule. A number of actuarial cost methods were endorsed by ERISA, permitting employers to meet the 30-year amortization schedule in a variety of ways, de- pending on their ability to pay and their need for tax deductions.

3. REPORTING AND DISCLOSURE Enforcement of funding and fiduciary standards without mean- ingful reporting and disclosure of accounting, investment, and actu- arial data would be nearly impossible. Understanding this, as well as the need for beneficiaries to have complete information about the nature of their plan and their entitlement to plan benefits, ERISA's framers detailed extensive reporting and disclosure re- quirements. Some of the provisions of the old WPPDA were merely carried forward in ERISA. However, ERISA framers believed that the new fiduciary, funding, and plan participation rules required new re- porting to the government and to participants. ERISA listed these requirements in great detail, leaving little flexibility on the part of the administrative agencies for deviation from its standards.

4. PLAN TERMINATION INSURANCE At the time of ERISA's passage, Congress believed that pension benefits in defined benefit plans could be guaranteed through per capita premiums paid by a pool of defined benefit plans. Two such risk pools were created: one for multiemployer pension plans and one for single employer pension plans. Contribution rates were es- tablished at $0.50 per capita for multiemployer plans and $1 per capita for single employer plans. ERISA's framers viewed the ter- mination insurance system as the insurer of last resort, hoping that ERISA's other provisions would limit its utilization.

5. MINIMUM PLAN PROVISIONS Congress recognized that millions of individuals working for em- ployers with pension plans had little hope of receiving benefits from such plans due to restrictive participation and service require- ments. One objective of ERISA was to increase benefit eligibility for those working for employers with plans. Congress sought to reach this objective by requiring all tax-quali- fied plans to meet certain eligibility, participation, and vesting rules. In addition, plans had to demonstrate that they met certain "nondiscriminative" standards, i.e., they were for the benefit of a broad segment of the employer's work force. These concepts were in direct conflict with the reasons why many employers-and some unions-had established pension plans in the first place: to reward full-career employees of the employers sponsoring the plans. The requirement to broaden participation to the entire work force of an employer made it more expensive to provide pension benefits to the favored employees. Earlier benefit eligibility (vesting) meant that shorter-service employees would re- ceive some benefit as well. Pension benefits became more of an en- titlement then a reward.

6. INCREASED PENSION COVERAGE ERISA's framers believed that the fact that a large number of employers in the private sector did not offer pension plans was a serious problem. The report of the Committee on Education and Labor (House Report No. 93-533, October 2, 1973) stated that a goal of ERISA was to "promote a renewed expansion of private retire- ment plans and increase the number of participants receiving pri- vate retirement benefits." The committee report, however, en- dorsed the concept that this increase in coverage would take place voluntarily through tax and other incentives that already existed at the time of ERISA's passage. C. WHAT ERISA DID NOT ATTEMPT Any law such as ERISA ultimately reflects the process of com- promise inherent in the legislative process. Some issues were not addressed due to their complexity or to effective lobbying by inter- est groups. Others were not addressed because the Congress in 1974 believed that its goals could be achieved through a completely vol- untary private pension system.

1. PORTABILITY CLEARINGHOUSE The 1965 President's Cabinet committee report called for a number of reforms including a national portability clearinghouse for employee pensions. That committee, as well as some promoters of ERISA, argued that some mechanism needed to exist for individ- uals to "take their benefits" with them when they changed employ- ers. In Congress, Senator Hartke and others argued that a national portability fund should be established with the requirement that employers would transfer assets equal to the accrued benefits of terminated employees to the fund. These efforts ultimately failed. Instead, Congress called for study of the portability issue and the establishment of a central record- keeping center for vested benefits of terminated employees within the Social Security Administration. A mandated study supported by funding from the Department of Labor found that while portability networks do exist among pen- sion plans, they are usually limited to plans that are linked by a common union or affiliated employers. The major exception to this rule is TIAA-CREF, covering college and university personnel. Neither the administrative agencies nor Congress formulated leg- islative proposals based on the report. However, in 1981, the Presi- dent's Commission on Pension Policy proposed the establishment of a Minimum Universal Pension System [MUPS] with a portability clearinghouse feature. Representative Claude Pepper introduced legislation later that year calling for the establishment of such a system. To date, that legislation has had little exposure. 2. PUBLc EMPLOYEE PLAN REGULATION Early versions of ERISA would have extended ERISA's provi- sions of Federal as well as State and local pension plans. The final legislation, passed on Labor Day 1974, called only for a congression- al study of these plans and the need for Federal legislation. The House Pension Task Force report issued in 1978 was the product of that effort. Congress was hampered by the lack of knowledge of State and local pension plan problems in 1974. In addition, lobbyists for State and local interests convinced many in Congress that the appropri- ate arena for reform was at the State and local level. By calling for a congressional study of public employee pension plans, Congress deferred a final decision on both the need for reform and the appropriate level of government from which to direct any such reform if needed.

3. MANDATORY EMPLOYEE PENSIONS The legislative record of ERISA is filled with endorsements of the belief that government should encourage the voluntary growth of employee pension plans. While ad hoc groups existed at the time to study and advocate the issue of mandatory employer-based pen- sions, that issue was not seriously discussed in Congress. When ERISA was passed, its supporters were optimistic about the prospect of an expanded employee pension system providing a supplement to Social Security for workers. In a sense, the establish- ment of the individual retirement account [IRA] was an attempt to extend pension coverage to workers employed at establishments without employee pension plans.

4. INFLATION PROTECTION While the issue of protecting the value of pensions against infla-, tion was discussed in Congress, ERISA established no guidelines or rules with regard to postretirement benefit increases. Instead, the already existing mechanics of ad hoc benefits increases either through unilateral employer action or through the collective proc- ess were to be relied upon. D. WHAT ERISA ACCOMPLISHED FOR PARTICIPANTS Partly as a result of ERISA and partly as a result of other events, employees are more aware of the importance of well-funded employee pension plans than they were at the time ERISA was en- acted. While ERISA's reporting and disclosure requirements have been the subject of criticism from many pension industry groups, it seems clear-particularly to plan participants-that many of those requirements have had a significant impact on plan participants. For example, when the President's Commission on Pension Policy conducted a national household survey on benefit entitlement, most individuals with pension entitlements referred to a plan de- scription booklet available in their home to answer the survey. Plan participants still complain that the summary plan descrip- tion booklets are difficult to understand. Part of this may be due to the complexities of the plans themselves. In the interest of legal correctness, however, these booklets are often written in such a way as to require personal counseling to understand their content. Many of the early complaints from the pension industry about the burdens of reporting and disclosure to the Federal Government have been alleviated as standardization and computerization have reduced compliance costs. Since private pension coverage has not changed significantly since ERISA's passage, the impact of ERISA's participation and vesting requirements on total benefit entitlement is difficult, if not impossible, to assess. Certainly, within pension plans, benefits are more equitably distributed then they were prior to ERISA. Howev- er, this very equity may have led to slower new plan formation. Improvements in vesting rules have certainly spread pension benefits among more individuals. Even so, preliminary evidence from the 1983 Current Population Survey indicates that the pro- portion of plan participants with vested benefit entitlements has remainded essentially unchanged since 1979. The funding and fiduciary standards of ERISA have probably had the biggest impact of any other provisions on plan partici- pants. ERISA's fiduciary rules now require trustees to oversee compe- tent administration of plans and prudent investment of plan assets. The endorsement of modern portfolio theory by the Department of Labor's prudence standards has required trustees and pension port- folio managers to increase their level of sophistication in pension fund investments. ERISA's funding standards, while ineffective under certain economic conditions (discussed later), have generally worked well to increase the benefit security of plan participants and beneficiaries. Most terminated pension plans are fully funded for vested benefits. Annual surveys conducted by Johnson & Hig- gins have shown that a significant portion of ongoing pension plans are also fully funded for vested benefits. The pension benefit guarantees established through the termina- tion insurance system administered by the PBGC has helped avoid another situation for plan participants like the Studebaker plan termination. Since benefits are frozen upon termination, however, there is growing awareness of the limited value of even these bene- fit guarantees. Nevertheless, for the level of guarantees covered by termination insurance, the system has worked well to protect the interests of participants and beneficiaries. This has occurred, however, only be- cause Congress has been willing to periodically adjust contribution rates upward, cut back on the level of benefits guaranteed, and in- crease employer liabilities. Whether this system can withstand future changes in economic conditions is unclear. E. WHAT ERISA FAILED TO DO FOR PARTICIPANTS

1. BENEFIT SECURITY Ironically, weaknesses in ERISA's provisions to protect pension benefits have emerged recently for quite opposite reasons. First, in 1982 the large number of business failures prompted by the reces- sion and structural shifts in the economy demonstrated the inad- equacy of the termination insurance premium levels. One business failure of a company like Chrysler would have totally disrupted the termination insurance system. This first problem is serious since it highlights a potential weakness in ERISA's funding standards and may point to future equity concerns with regard to plan termina- tion insurance. Second, the high interest rates earned on pension asset invest- ments in recent years have created incentives for many companies to terminate pension plans and recover funds that are determined to be "excess assets" according to ERISA definitions.

(A) FUNDING The passage of ERISA led to substantial changes in the way pen- sion plans are funded. ERISA calls for companies to finance de- fined benefit pension plans by paying "normal costs" as they are incurred each year. However, plan sponsors may amortize or spread so-called "supplemental costs" or past service costs over a number of years. Simply put, benefit pension plans must pay annu- ally for each year's current costs but they may spread payments for past service credits and any benefit improvements over a 30- year (or 40-year) period. The 30-year amortization is significant for two reasons. First, vir- tually all private plans are not indexed to inflation. Workers are dependent on ad hoc benefit improvements to maintain the real value of their pension benefits. Nearly all of these ad hoc inflation adjustments will be paid for over a 30-year period. Second, as both workers and employers realize the inadequacy and the uncertainty surrounding the Social Security system, they are seeking a greater role for private pension benefits. Again, all of these benefit im- provements are likely to be paid for over a 30-year period. Why is this important? Take the example of a 35-year-old worker who plans to retire 30 years from now at the age of 65. We know today at least part of the cost of every benefit improvement made in his plan between now and his retirement date (and beyond, if he is lucky) will be borne by his employer and future generations of workers after he retires. In other words, even though ERISA calls for private pensions to be "funded," ERISA still permits an ele- ment of intergenerational income transfer to exist, not unlike that which we find in Social Security. Even though all of this is perfectly legal and acceptable by ERISA's actuarial standards, this may lead to serious underfund- ing, particularly in collectively-bargained plans of the flat benefit and unit benefit type.

(B) PLAN TERMINATION INSURANCE The plan termination insurance program administered by the Pension Benefit Guarantee Corporation [PBGC] is made up of two risk pools (one comprised of multiemployer plans and one of single employer plans) with the benefit guarantees financed by per capita ''assessments," "taxes," or "contributions" on the individual plans. Under this system, the well-funded plans subsidize the plans that terminate and are underfunded. Since its inception the PBGC has explored the possibility of de- veloping a so-called "risk-related" premium to eliminate the obvi- ous inequities that exist in the current system. These efforts have failed to produce such a premium structure due to the complexities of the problem, the inadequacy of available data to determine risk, and the judgment that with premiums at past or even current levels the inequities were not too serious. The complexities certainly are there. Not only does one need to assess the funding status of the plan, but one needs to analyze the probability of the business failure of the plan sponsor-the employ- er. Due to the 1978 reforms instituted by the DOL, it is at least pos- sible to develop a variable premium that would have two compo- nents: (1) A low per capita component, and (2) a variable compo- nent that would be a function of unfunded vested benefits (current- ly disclosed on schedule B). Some form of "risk-related" or "exposure-related" premium may need to be instituted as PBGC premium levels rise in the future. Companies would be rewarded for funding their plans, not pun- ished. It seems unrealistic that pension benefits can be truly insured with per capita taxes of only several dollars a year. Experience since 1974 has shown that plan sponsors will utilize the termina- tion insurance system much more heavily than anticipated by ERISA's designers. While these designers may have wished that the 30-year amortization rule would eventually eliminate unfunded benefit liabilities in private pension plans, this has not been the case.

(C) THE EXCESS ASSETS ISSUE Economic conditions combined in the past few years to provide incentives for companies to terminate their defined benefit pension plans and recover millions of dollars of so-called excess assets. Well over a billion dollars in assets are currently in the process of being recaptured by corporate plan sponsors. High interest rates on investment earnings have permitted many pension plans to find themselves "ahead" of ERISA's minimum funding schedule. ERISA's framers could not have anticipated the large disparity between long-term interest rates used by actuaries and actual earnings on invested funds experienced by pension plans in the past few years. Last year, for example, the Standard & Poor's 500 index, an index used by many pension funds to judge their equity performance, increased by 22.4 percent. A typical pen- sion fund might use a 7 to 8 percent long-term interest assumption. This difference can lead to between 20 and 30 percent difference in the estimated pension fund liabilities in many plans. In addition, benefit obligations of terminating plans may be substantially less than in continuing plans. These two factors have combined to create a substantial economic incentive for companies to terminate their defined benefit plans. While ERISA would also permit em- ployers to reduce future contributions until the "excess" is elimi- nated, the attraction of a "lump sum" recovery has been too great for many to resist. By recovering assets that are in excess of the benefit obligations of a terminating plan, beneficiaries already have lost hundreds of millions of dollars in potential retirement benefits from their plans. Furthermore, their pension benefits may be forever frozen. In addition, if economic conditions were to change, the so-called ''excess" could very well disappear. For those participants in final salary plans, the value of their benefits earned to date may be substantially reduced even if a guaranteed annuity is purchased on their behalf for past service since their benefits under the annuity arrangement would be com- puted on current salary. This loss of benefits could exist even in the case of employers who establish successor plans. 2. COVERAGE Estimates indicate that approximately half of the private work force currently participates in at least one of the half million exist- ing private pension plans. About 85 percent of the State and local workers are covered by pension plans, and virtually all Federal personnel are covered by one of the Federal pension systems. Two companion surveys were conducted on the issue of pension coverage in the United States in 1979. One was conducted by the Census Bureau as a supplement to the Current Population Survey and the other was conducted by the President's Commission on Pension Policy. Table 2 shows the results of these two surveys for the private sector. The percentage of "coverage" depends in part on which definition of coverage one wished to use and how much of the work force one wants to observe. For this table, an individual is considered covered only if he or she actually is a participant in a pension plan. Some pension industry groups define coverage as being employed at a firm or organization that has a tax-qualified pension plan.

TABLE 2.-GROWTH INRETIREMENT COVERAGE [In percent]

Coverageforspecific labor forces 1940 1950 1960 1970 1975 1977

SocialSecurity coverage of all paidemployment...... 1 55 65 88 87 90 90 Federalemployee retirement coverageof all Federal employment...... 2100 2 100 2 100 2 100 2 100 2 100 Stateand local retirement coverageof all Stateand localemployment...... (3) (2) 4 82 5 83 (2) 87 Private pensioncoverage of nongovernment wageand salary nonagricultural employment...... 14.5 25.0 40.8 45.1 48.6 (2) 1For year 1939. 2Approimate. ' Not available. 4 Foryear 1962. 5 Foryear 1972. SourceEconomic Report of thePresident, Janary 1980,table B, p. 34.Social Security Bulletin, Annual Statistical Supplement, 1976, table 7. SpecialAnalysis of theFederal Budget, 1981, table 1-4, p. 288.Statistical Abstract, 1979, table 543, p. 339. Social Security Bulletin, November 1977,p. 27. In its deliberations and final report, the President's Commission on Pension Policy used participation as a definition of coverage. The other issue to address is which portion of the labor force to in- clude in the numerator and denominator of the coverage fraction. The President's Commission chose to use the participation figures for all workers over the age of 25 as a benchmark of coverage. For the private sector, that yields 47 percent based on the Pension Commission's survey and 52 percent based on the Census survey. Another definition sometimes used is participation in plans for those meeting so-called ERISA standards (age 25, full time, more than 1 year at current employer). While useful for certain indus- tries, the Pension Commission rejected this for all industries since approximately half of the work force over the age of 25 was ex- cluded from the analysis. The ultimate question is who will eventually be covered for a substantial number of years in a pension plan, not how many are covered at any one point in time. Pension Commission labor force forecasting models indicated in 1981 that pension plan participa- tion could not be expected to increase significantly in upcoming years without a substantial change in tax policies and availability of employer-based pension plans. The latter was seen as unlikely since forecasts of structural shifts in the economy seemed to indi- cate employer-based pension coverage might decline rather than in- crease. The Employee Benefit Research Institute (EBRI) used dif- ferent assumptions with the Pension Commission model and fore- casted increases in pension coverage in upcoming years. Use of these models is problematic for policymakers due to the sensitivity of the results to the models' assumptions. For example, these models assign future coverage based on projections of age, sex, income, occupation, industry affiliation, and union affiliation of the workers. An increase of 1 percent in the wage growth as- sumption leads to an increase of 5 percent in the labor force pen- sion coverage figures. A more fruitful approach might be to look carefully at future trends in industry structure: where tomorrow's jobs will be. Prelim- inary findings from the 1983 Current Population Survey indicate that the recent recession may have accelerated the structural shifts in employment predicted by the President's Commission, leading to actual reductions in the proportion of the labor force covered by pension plans. Using a slightly different definition of coverage than the Presi- dent's Commission, EBRI has found that pension coverage actually decreased approximately 4 percent for all civilian workers between 1979 and 1983.2 Since public sector civilian employees did not expe- rience a decline in coverage, it can be assumed that pension cover- age for private sector employees declined by more than 4 percent. While some improvement in these figures might already be real- ized as the economy, and employment, improves, a long-term shift away from high-pension-coverage industry to historically low-pen- sion-coverage service sectors appears to be underway. While ERISA should not be blamed for these declines, it appears that the com- bination of ERISA's provisions has not led to the increased cover- age as ERISA's founders had hoped for.

2 Employee Benefit Research Institute. Issue Brief No. 33. New Survey Findings on Pension Coverage and Benefit Entitlement. Washington, EBRI, 1984. On the other hand, according to tabulations by EBRI of the 1983 CPS, the proportion of those within pension plans who are vested has increased since 1979. This trend should be expected to increase at a moderate pace as the baby-boom generation ages and earns service credits under pension plans that are in place. Why ERISA's framers believed that more complete regulation of a totally voluntary pension system would lead to greater coverage is not at all clear. It seems only logical that one of the consequenc- es of the funding, fiduciary, and benefit eligibility rules would be that more people within any given plan would likely become eligi- ble for greater benefits. Even without needless administrative bur- dens, these plans would necessarily become more expensive. More expensive plans would be less attractive to many employers, slow- ing the growth of new plans. The evidence on pension coverage to date seems to indicate that in the private sector, a plateau has been reached on pension cover- age. As long as pension plan establishment and termination remain voluntary acts, it seems unreasonable to expect substantial improvements in the proportion of the private labor force covered by these plans.

3. PUBLIC PENSION PLAN REGULATION Since the issuance of the House Pension Task Force report in 1978, some form of Federal legislation governing State and local pension plans has almost continuously been before the Congress. In the 97th Congress, H.R. 4928 and H.R. 4929, the Public Employee Retirement Income Security Act [PERISA], were introduced by Representatives Phillip Burton, chairman of the House Labor-Man- agement Subcommittee, and John Erlenborn, ranking minority member of the Education and Labor Committee. Identical meas- ures (S. 2105 and S. 2106) were introduced in the Senate in Febru- ary by Senator John Chafee, chairman of the Finance Subcommit- tee on Savings, Pensions and Investment Policy. The President's Commission on Pension Policy supported the legislation, though it believed that Federal regulation of funding and benefit standards should also be included in any such bill. After State and local government officials raised certain objec- tions at hearings on the bills in February 1983, the legislation was rewritten with several "deregulatory" provisions. The new legisla- tion, called the Public Employee Pension Plan Reporting and Ac- countability Act [PEPPRA], was reported out of the House Educa- tion and Labor Committee with only one dissenting vote. As rewritten, PEPPRA would require reporting and disclosure of certain benefit, financial and actuarial information, as well as es- tablish fiduciary standards and enforcement procedures. States are granted an exemption from certain provisions if the State's Gover- nor certifies that State laws have "substantially equivalent" provi- sions. Between 1979 and 1981, the President's Commission on Pension Policy reviewed the findings of the House Pension Task Force report, initiated and coordinated new research on State and local pension plans, and held hearings around the country on problems with these plans. The final report of the President's Commission agreed with the House Pension Task Force that problems exist in the following areas: participation, vesting, reporting, disclosure, funding standards, fiduciary responsibility, limits on benefits or contributions, survivor benefits, and plan termination insurance. Ironically, inadequate reporting and disclosure have hampered the development of conclusive research in some of these areas. However, enough is now known about some problem areas to sug- gest the need for immediate reform.

(A) INADEQUATE FIDUCIARY STANDARDS Prior to the establishment of ERISA, private sector employees had to rely on State and local laws to protect them from abuse by plan administrators and trustees. Pension experts universally agree that the establishment of uniform fiduciary standards by ERISA has had a major influence on ending pension fund misman- agement. Both the President's Commission and the House Pension Task Force concluded that public employees need this same protec- tion. The absence of uniform fiduciary standards has led to abuses such as conflicts of interest in management, and unprofessional in- vestment practices. In the words of the Pension Task Force report: There is virtual unanimity within the pension communi- ty that those who have control of pension assets should be held to high standards of behavior and should face liability upon failing to satisfy that standard * * * throughout the universe of State and local government retirement systems there is a virtual absence of clear guidelines in this vital area. A study conducted by Louis Kohlmeier for the Twentieth Centu- ry Fund documents widespread conflicts of interest in the manage- ment of State and local pension funds: One of the most persistent conflict-of-interest situations in the management of public pension funds results from the policy, followed by many plans, of hiring local bankers, brokers and investment advisors and the practice of invest- ing in local securities, even though better or lower cost services and higher yielding investments may well be available outside local boundaries. Some of this activity is well-intentioned: legislators and plan ad- ministrators sometimes seek to encourage local business. Often, State law will specify that certain types of investments, such as mortgages or municipal bonds, must make up a fixed portion of the pension portfolio. In addition, some State laws exclude investments in certain financial instruments such as corporate stocks. Whether or not well-intentioned, any sacrifice in investment return due to these restrictions may not be in the interest of either the beneficiaries of the pension plans or the taxpayers that support them. Another area of fiduciary abuse highlighted by both the Pension Task Force report and the Kohlmeier study concerns the absence in many States of professional investment management. Frequent- ly, pension fund trustees are nonexpert in the field of portfolio management. This "often produces investment policies and prac- tices that are significantly less valuable than that expected from professional investment advisors and managers, and generally found in private sector plans." In spite of the number of reports calling for changes in State and local fiduciary practices, local reform has been extremely slow, and the prospects for significant changes in the near future seem remote. Even the experience in States that have made moderate progress in pension reform illustrate the dangers of relying on that process for significant change. The battle for reform of California's fiduci- ary, investment management, and reporting and disclosure prac- tices provides a current example. For the past several years, the State legislature has been debat- ing reform of the management practices of California's two large public employee pension funds, totaling over $30 billion in assets. A Joint Committee on Public Pension Fund Investment has, for the past 2 years, hired consultants, held hearings, conducted studies, sought the advice of experts throughout the country, and drafted legislation. Consultant reports to the committee found that the funds were difficult to oversee due to inadequate reporting and disclosure, that fund administration was not insulated from the political process, that portfolio performance suffered from the quality and quantity of resources devoted to investment staff, and that guidelines for trustee behavior did not exist. The major consultant to the commit- tee, Dr. Mercy Avrin, reported that the taxpayers of California could save hundreds of millions of dollars per year by increasing the return of the pension portfolio to reasonable levels. The bills that were presented to the State legislature by the joint committee in 1982 and 1983 have been nearly universally praised by pension and investment experts. At first, it seemed successful passage of these bills was ensured. In 1982, an important break- through was made with the enactment of an ERISA-like prudence standard. However, when the legislature attempted to add teeth to this provision, with passage of a bill adding a comprehensive re- porting and disclosure provision and separating the investment board from executive branch and legislative influence, the Gover- nor vetoed the bill. Even if the legislature eventually prevails, this veto illustrates the difficulty of significant pension reform on the State and local level. It is unlikely that most States and munici- palities will devote the time and resources for reform that Califor- nia has. By their very size relative to the budgets of their governmental sponsors, public pension funds are easy targets for budgetary and political manipulation. Politicians are unlikely to relinquish their control over these funds voluntarily. Furthermore, full disclosure may lead to embarrassing reports of underperformance by political appointees. As long as the disclosure of this performance can be hidden or delayed, those responsible will not be held accountable. (B) INADEQUATE REPORTING AND DISCLOSURE Virtually every major study of State and local reporting prac- tices has found serious inadequacies. Frequently, important finan- cial, actuarial, and accounting calculations are either not per- formed or not revealed. In many instances, plan participants are not even informed of their basic plan benefits and legal rights through simple summary plan descriptions. Most experts agree that complete reporting and disclosure of fi- nancial and benefit information is the least intrusive way to reduce abuse by pension trustees and plan administrators. Due to the highly complex nature of pensions, inadequate disclosure makes it impossible for even experts to detect abuse or mismanagement until it is too late: when pension promises are broken or additional taxes must be raised to prevent insolvency. This point was emphasized by Louis Kohlmeier in his study of asset management practices: Most public pension plans make financial reports of some kind to the legislature, to the Governor or mayor, to employees and/or the general public. The great majority of such disclosures are wholly inadequate to allow legislators, employees or the public to judge the inadequacy of fund administration.* * * Rarely do reports disclose [invest- ment information capable of being analyzed]. The House Pension Task Force concurred when it concluded that the "potential for abuse is great due to the lack of independent and external reviews of the operations of many plans." Late in 1978, the newly established President's Commission on Pension Policy began to coordinate an interagency research effort on state and local pension plans that resulted in three major re- ports in 1980 and 1981. The first report, conducted by the Urban Institute, examined a sample of 100 large pension plans. While these plans are generally considered to have the best reporting and disclosure of all state and local plans, table 3 shows that even they have serious gaps in disclosure.

TABLE 3.-FREQUENCY OF DISCLOSURE OF PARTICULAR ITEMS INANNUAL REPORTS

Percentageof Item plansincluding

Auditors' opinion...... 40 Reportof assets,liabilities, etc ...... 99 Statem ent of changesin reserves ...... 54 Statement of factors (e.g.,litigation andtrends) that mayaffect financing andoperation...... 33 Statement of investment policies andrestrictions...... 47 Portfolio byasset type ...... 93 Fundingpolicy for employers andm em bers ...... 58 Dateof last actuarial report ...... 65 Changesin actuarial assumptions ...... 23 Summary of actuarial assum ptions ...... 37 Amount of liability (actuarial balancesheet) ...... 71 Numberof former employeesvested but notyet getting benefits...... 43 Number of beneficiaries ...... 86 Annual reports should contain complete accounting, actuarial, and financial information. However, the federally sponsored Urban Institute study showed a number of deficiencies. Only 40 percent contained an auditor's opinion, and only 33 percent contained a statement of factors that might affect financing and operation. While 99 percent contained a statement of assets and liabilities and 71 percent contained an actuarial balance sheet, only 37 percent disclosed actuarial assumptions used to perform the calculations, and only 23 percent disclosed changes in actuarial assumptions that might affect year-to-year variations in the reported numbers. Without these further disclosures, the other figures are virtually meaningless to experts. Disclosure of adequate investment criteria and performance was also found lacking: only 47 percent disclosed even a statement of investment policies and restrictions. The second product of the Federal pension research effort was a report issued in 1981 by SRI International on small- and medium- sized state and local pension plans. Table 4 shows a summary of SRI's reporting and disclosure findings.

TABLE 4.-AVAILABILITY OF INFORMATION ON PLANS (Inpercent]

Document Document Typeof document Documentexists availableto availableto the employee public

Descriptive booklet for participants...... 58.0 49.3 40.7 Law , ordinancesor code ...... 71.3 66.7 63.3 Recentannual report ...... 92.0 84.7 77.3 Recentactuarial report ...... 88.0 77.3 70.0 IRS550G form...... 92.7 83.3 76.0

According to the report, fewer than half the employees have a booklet describing the benefits and eligibility criteria for their plans. Other documents such as annual reports, actuarial reports, and ordinances governing the plans are too often available to nei- ther the employees nor the public. The third report of the Federal research effort was a report on financial reporting and disclosure prepared by the Municipal Fi- nance Officers Association. They concluded that "available infor- mation indicates such reporting is today inadequate and confused and clearly in need of repair. Several factors contribute to the lack of good disclosure about [State and local] pension systems.* * * Further, the MFOA concluded that these deficiencies were due to a "lack of general authoritative standards for system disclosure and of enforcement of such standards that do exist."

(C) INADEQUATE FUNDING STANDARDS The reporting and disclosure evidence that is available indicates that a potentially serious problem exists with regard to inadequate funding of state and local pension plans. Whether the failure to disclose funding policies means that none exist or that the plans may fall into insolvency is difficult to discern.

37-569 0 - 84 - 4 Part of the Federal research effort referred to earlier involved an attempt to estimate the funding status of State and local pension plans. The findings of this effort show a mixed picture: (1) Large plans, in the aggregate, appear reasonably well funded. (2) Some large plans face funding problems. (3) Small plans cannot be easily evaluated due to inadequate reporting and disclosure. (4) Small plans appear very vulnerable since many are de- pendent on outside sources of funds for their annual contribu- tions. While the Federal research projects do not provide conclusive proof of a national underfunding problem for State and local pen- sion plans, they do offer evidence of pension plan vulnerability to changes in benefit policies, interest earnings, contribution sources, State and Federal policies, State and Federal budgets, and many other factors that are likely to affect them. Better reporting and disclosure would permit future research efforts to determine whether the funding problems faced by some pension plans are suf- ficiently widespread to warrant Federal standards. High interest rates in the past few years have temporarily im- proved the funding status of some plans since the Federal studies were conducted. However, the procedures and standards for fund- ing remain largely the same, indicating a serious potential for ben- efit loss or taxpayer burden in future years. F. ARE ERISA'S GOALS IN CONFLICT WITH FEDERAL TAX POLICY? ERISA endorsed the concept of group pension plans through em- ployer and employee contributions from salary. The nondiscrimina- tion rules, participation, and eligibility rules were designed to ensure that a broad cross-section of workers would be covered by pension plans. At the same time, ERISA called for the establishment of IRA's for those not working for employers with tax-qualified plans. While the intent appears to have been to extend favorable tax treatment for retirement saving to those not covered by pensions, current rules governing IRA's and their income deferral programs appear to have touched off the rapid growth of tax-favored programs that potentially undermine many of ERISA's principles. For one, IRA's Keogh plans, and now section 401(k) plans are be- ginning to be promoted as tax shelters rather than retirement plans. This past tax season, banks and others explicitly marketed IRA's as tax shelters, even informing individuals in high tax brack- ets how long they needed to keep their money in these programs and still come out ahead of other investments even with the tax penalties for early withdrawal. Table 5, from an article in the Wall Street Journal, shows one such illustration. 43 TABLE 5.-IRA's AREN'T FOREVER [Yearsrequired for an IRA to yieldmore, after penalty, than a taxableinvestment]

Pretaxinterest rate (percent) Marginaltax rate(percent) 8 10 15

50 ...... 6 5 4 40 ...... 7 5 4 30 ...... 7 6 4

Source:Wall Street Journal, "Investing in an IRADoesn't Have to MeanTying Up Your Money Until You'ne 59%," Monday,Mar. 26,1984, p. 31. IRA and Keogh utilization is much less equally distributed among income groups than are regular tax-qualified group plans. Table 6 show preliminary tabulations by EBRI from the 1983 Cur- rent Population Survey that illustrate this point. While group pen- sion plans are also utilized more by higher income groups, IRA dis- tributions are much more extreme. Without the benefit of ERISA's nondiscrimination and other rules, the further development of these programs may encourage employers to substitute them for regular group plans.

TABLE 6.-PARTICIPATION INALTERNATIVE RETIREMENT INCOME PROGRAMS BY EARNINGS FOR 1982 (Civilianemployment, May1983]

Participantswithin earnings groups (percent) Earnings IRA 401(k) Pensionplan

Total...... 16 .9 2.7 43.1 $1to $4,999 ...... 7.1 0.2 7.8 $5,000 to $9,999 ...... 8.5 0.8 23.5 $10,000 to $14,999...... 11.1 2.1 47.2 $15,000 to $19,999...... 17.3 3.5 63.7 $20,000 to $24,999...... 20.1 4.5 72.8 $25,000 to $29,999 ...... 28.4 6.2 73.0 $30,000 to $49,999...... 38.7 8.2 77.4 $50,000 andover ...... 57.6 8.6 74.4

Note-Numbers and percents may not addto totals dueto roundingand exclusion of respondentswhose earnings were not reported. Source:Preliminary Employee Benefit Research Institute tabulations of the May1983 EBRI/HHS Current Population Survey (CPS) Pension Supplement. EBRI tabulations also show that a majority, 53 percent, of those utilizing IRA's did so as a tax shelter, while only 38 percent did so for retirement savings.3 G. CONCLUSION ERISA has had a major impact on the management of pension plans throughout the Nation. Its goals of reforming the funding methods, investment practices, and administration of private pen- sion plans have largely been achieved. Ten years of experience with ERISA, however, has shown that certain flaws do exist in the law. Protection of pension benefits has not been fully realized by ERISA's combination of funding, fiduci-

3 Employee Benefit Research Institute. Issue Brief No. 32. Individual Retirement Accounts: Characteristics and Policy Implications. Washington, EBRI, 1984. ary, reporting and disclosure and termination insurance provisions. State and local governments have been reluctant to reform pension systems covering their employees. Large segments of the private labor force remain uncovered by pension plans altogether. Perhaps most important, Federal tax policy seems to be under- mining the very principles on which ERISA was based: the encour- agement of broad-based pension protection through group employ- er-sponsored plans. Beginning with ERISA's passage, Federal tax policy has encouraged individual savings. Some employers have ter- minated broad-based group plans and have established these plans on voluntary individual participation. Research has consistently shown that this tax-incentive approach to encourage individual retirement savings leads to relatively high participation by high-income individuals and relatively low partici- pation by low-income individuals. Further encouragement of these programs by Congress will, no doubt, lead to further erosion of pri- vate pension coverage among moderate income workers. It seems appropriate, 10 years later, for Congress to rethink its rejection of some form of minimum pension plan and some form of portability clearinghouse for private sector employees. The momen- tum for tax-induced savings appears, at this juncture, unstoppable. If, however, the retirement income needs of moderate income couples and individuals is not met in part by funded private pen- sions in the future, these Federal tax policies may lead to either a highly unequal post-retirement income distribution or to excessive demands on the Social Security system. H. BIBLIOGRAPHY Employee Benefit Research Institute. Issue Brief No. 32. Individual Retirement Accounts: Characteristics and Policy Implications. Washington, EBRI, 1984. --. Issue Brief No. 33. New Survey Findings on Pension Cover- age and Benefit Entitlement. Washington, EBRI, 1984. ICF, Inc. Prepared for the American Council of Life Insurance. Pension Coverage and Expected Retirement Benefits. Washing- ton, ICF, 1982. Peterson, John. Public Pension System Financial Disclosure. Mu- nicipal Finance Officers Association, 1980. President's Commission on Pension Policy. Final Report. Coming of Age: Toward a Retirement Income Policy. Washington, U.S. Govt. Print. Off., 1981. --. Technical Appendix. Coming of Age: Toward a Retirement Income Policy. Washington, U.S. Govt. Print. Off., 1981. The appendix represents a compilation of the findings of nearly 50 research projects conducted between 1979 and 1981 by the Commission and numerous Federal agencies. Twentieth Century Fund. Conflicts of Interest: State and Local Pension Fund Asset Management. Twentieth Century Fund, 1976. Woodruff, Thomas C. Dollars and Sense: The Case for State and Local Pension Reform. Washington, American Federation of State, County, and Municipal Employees, 1984. Chapter 3 POST-ERISA LEGISLATION (Prepared by Dan M. McGill, Chairman and Research Director, Pension Research Council, Wharton School of the University of Pennsylvania) A. INTRODUCTION An extensive body of pension legislation has been enacted by Congress since ERISA. Indeed, there has been some such legisla- tion in each of the intervening years except 1977, 1979, and 1983. Some of the enactments were of limited scope but others were major pieces of legislation, of far-reaching import. In most cases, what is here characterized as pension legislation took the form of some pension-related provisions in tax legislation of much broader scope and objective. In this chapter, these pension measures will be considered chronologically rather than topically, even though this approach leaves something to be desired pedagogically.' Because of space limitations, only the central features of the successive pieces of legislation can be described. During the last 10 years, the Federal agencies responsible for ad- ministering and enforcing ERISA-principally the Internal Reve- nue Service, the Department of Labor, and the Pension Benefit Guaranty Corporation-have issued hundreds of regulations, rul- ings, guidelines and interpretations. Collectively, these have cre- ated a significant body of substantive law, but they are not the sub- ject matter of this chapter. The same is true of judicial decisions.

B. TAX REDUCTION ACT OF 1975 2 The Tax Reduction Act of 1975 was a significant revenue meas- ure but it contained only two sections dealing with tax-favored asset accumulation plans.

1. EMPLOYEE STOCK OWNERSHIP PLANS The term employee stock ownership plan [ESOP] has been used to describe a wide variety of arrangements. In a broad sense, an employee stock ownership plan can be defined as any type of de- fined contribution plan that invests some or all of its assets in em- ployer securities. It may be qualified or nonqualified as far as tax treatment is concerned. Nonqualified plans are generally main- tained for the benefit of top management. This chapter is con-

' For a topical analysis of the subject matter of these statutory enactments, see Dan M. McGill, Fundamentals of Private Pensions, 5th Ed., Richard D. Irwin, Inc., Homewood, IL, 1984. 2 Public Law 94-12. cerned only with ESOP's that either are tax-qualified or are de- signed to benefit the employees generally. Employee benefit plans that hold employer securities in trust for eligible employee participants have long been encouraged through favorable tax treatment as a device for giving the labor force a broader stake in the private enterprise system and thereby enhanc- ing productivity. Employer contributions to qualified ESOP's have always been deductible within specified limits and the plan and its participants have enjoyed the usual perquisites of a qualified plan. Employer contributions are in the form of qualifying securities or cash applied to the purchase of employer securities. Small, closely held corporations have found the ESOP especially attractive, since it could be used to further broad corporate pur- poses as well as the special interests of the owners. For example, a "leveraged" ESOP can be used to raise capital for the sponsoring firm. Under a leveraged ESOP, the trustee of the trust created under the plan arranges a loan from a lending institution and uses the loan proceeds to purchase employer stock. The stock so ac- quired is held by the trustee and gradually allocated to the partici- pants as cash contributions are made on their behalf to the plan. The stock is pledged as collateral for the loan which is usually guaranteed by the employer or other party. An ESOP can also be used to provide estate liquidity for the principal owners by serving as a "captive" market for the stock (stock valuation procedures must be fair and meet certain standards). In response to the strong interest of certain members of Congress in the ESOP, Congress included a provision in the Tax Reduction Act of 1975 designed to give a special impetus to the formation and maintenance of ESOP's by firms of all sizes. The act provided a tax credit (in lieu of a tax deduction) for ESOP's that met certain standards. The credit was expressed as a percentage of the capital investment of the firm for the tax year in question and was in ad- dition to the tax credit of up to 10 percent for the capital invest- ment itself. The ESOP tax credit was 1 percent of the new capital investment or, if lesser, the amount of contribution to the plan. An ESOP that qualified for this investment-related tax credit was usually referred to as a TRASOP [Tax Reduction Act Stock Ownership Plan]. If the employer contribution to a TRASOP did not exceed the tax credit, the arrangement was without cost to the employer (except possibly for the expense of administration, which is reimburseable to some extent).

2. TIMING OF CONTRIBUTIONS To KEOGH PLANS The Tax Reduction Act of 1975 contained a provision relating to the timing of contributions to retirement plans for self-employed persons. It has long been the general rule that an accrual basis emloyer can make deductible contributions for a taxable year to a qualified retirement plan at any time not later than the due date of his tax return, including extensions, for that year. Prior to ERISA, this privilege was not accorded to cash basis employers who had to make their contributions before the end of the taxable year in- volved. Primarily for the benefit of self-employed persons, ERISA extended to cash basis employers the privilege of deducting contri- butions to a qualified plan after the end of the taxable year, up to the due date of the tax return, including extensions. However, be- cause of certain effective date provisions, many self-employed per- sons would not be able to take advantage of the newly-granted privilege for several years. To correct this "deficiency," the Tax Reduction Act of 1975 pro- vided that a self-employed person could treat contributions made on or before the due date of his return, including extensions, as if they had been made on the last day of the taxable year to which the return relates. The provision was made applicable to plans in existence on January 1, 1974, as well as to those established there- after, but was operative only for plan years beginning after Decem- ber 31, 1974.

C. TAX REFORM ACT OF 1976 3 To encourage personal thrift and to provide an opportunity to persons not participating in tax-favored retirement programs to make their own individual arrangements for retirement income on a tax-favored basis, ERISA authorized the establishment of individ- ual retirement plans, now generally known as IRA's [individual re- tirement accounts]. Eligible persons could contribute up to the greater of 15 percent of compensation or $1,500 per year to such a plan and deduct the amount for Federal income tax purposes. Only employees and self-employed persons were eligible to establish these plans. The Tax Reform Act of 1976 liberalized the law to permit a married person with employment income to set aside an additional $250 per year in a subaccount for a spouse not working outside the home. Conversely, the employed person could contrib- ute up to $875 per year to his or her own IRA account and up to $875 to a separate IRA for the spouse. In either case, the total de- duction was to be limited to 15 percent of compensation. As will be seen this was the first of a series of legislative liberalizations of the IRA approach. As a further encouragement to TRASOP's the act permitted the sponsoring employer to take an additional tax credit for matching employee contributions to a TRASOP. This additional tax credit was limited to one-half percent of the employer's capital invest- ment for the year. The act contained a number of other provisions designed to en- courage individual retirement programs. One allowed amounts con- tributed by certain tax-exempt employers to buy shares of a closed end mutual fund to provide employee retirement benefits to be treated as contributions for the purchase of tax-deferred annuities. Another permitted an owner-employee to base his payments under a level premium deferred annuity on the average of the amounts that he could have contributed each year had the plan been in ex- istence during the preceding 3-year period. Finally, Armed Forces reservists and voluntary firefighters were made eligible to pur- chase IRA's subject to certain conditions and the Joint Committee

3 Public Law 94-455. on Taxation was instructed to study the feasibility of making more persons eligible for IRA's. To clear up the uncertainty over life insurer guaranteed income contracts [GIC's] and contractual arrangements not contemplating the providing of annuities for retiring employees, the act explicitly authorized a qualified pension plan to invest in an insurance com- pany contract utilizing a segregated asset account and not provid- ing for the payment of annuities.

D. REVENUE ACT OF 1978 4 1. SEP-IRA Following ERISA's authorization of individual retirement ac- counts, with deductibility of contributions up to the lesser of $1,500 per year or 15 percent.of compensation, many small employers lacking a formal retirement plan began making contributions up to the maximum permissible amount to the individual retirement ac- counts of some or all of their employees. In some cases, this ar- rangement involved nothing more than the convenience of invest- ment through payroll deduction, but in other cases the employer contribution represented an increase in salary or wages. The ar- rangement was informal and could operate in a manner that would be regarded by some as discriminatory. In considering approaches that would encourage small employers to establish formal retirement plans for their employees and there- by reduce the gaps in pension coverage of the private sector labor force, Congress recognized the potential in the use of employer-fi- nanced IRA's that had developed informally in the small business community. Building on and strengthening this approach, Congress included a provision in the Revenue Act of 1978 authorizing a sim- plified form of tax-qualified pension plan that would operate in a nondiscriminatory manner and would enable small employers to realize the advantages of a conventional tax-favored retirement plan without its complexity and administrative expense. The in- strument created by this act is called a SEP [simplified employee pension]. Since a SEP is in essence an arrangement under which the employer establishes and finances an IRA for each of his eligi- ble employees, it is commonly referred to as a SEP-IRA. In order to make the arrangement more comparable to a conven- tional pension plan and, hence, more attractive to employers, espe- cially small business firms, Congress raised to $7,500 (or, if lesser, 15 percent of compensation) the tax-deductible amount that could be contributed to an IRA that is part of a SEP meeting certain standards. For the increased limit to be available, the contributions must be made pursuant to a written allocation formula that is not discriminatory in favor of employees who are officers, shareholders, self-employed, or highly compensated. In addition to that general prohibition, contributions must be made on behalf of all employees who have attained age 25 and have performed service for the em- ployer for the current calendar year and 3 of the preceding 5 years. All contributions must be fully vested and cannot be subjected to

4 Public Law 95-600. employer restrictions on the employee's right to withdraw amounts in his individual account or annuity. SEP-IRA's are subject to the usual rules for individual retire- ment accounts or annuities. If the employer contribution in a par- ticular year is less than the amount the employee could have de- ducted, the employee is permitted to make up the difference.

2. DISTRIBUTIONS OF INVESTMENT COMPANY STOCK FROM SECTION 403(b) CUSTODIAL ACCOUNTS Prior to this legislation, shares of stock in regulated investment companies held in custodial accounts maintained pursuant to IRC section 403(b) could be distributed at any time to the employee. Concerned that the purpose of this tax-favored retirement arrange- ment could be frustrated by premature distributions, the Treasury Department proposed regulations that would have prohibited dis- tributions of such stock before the employee attained age 65 except for the employee's death, total disability or separation from service after age 55. Superseding the proposed regulation, this legislation authorized distributions of stock of a regulated investment company upon an employee's'death, disability, separation from service at any age, at- tainment of age 591/2 (whether or not separated from service), or encountering of financial hardship. Under the "financial hardship" rule, distributions were to be allowed if the hardship would satisfy the "hardship" test of qualified profit-sharing plans. It was contem- plated that the hardship rule would be interpreted in such a manner as to permit distributions to an employee for the purpose of purchasing a residence or meeting the higher education needs of the employee's children.

3. LIFE INSURANCE PENSION RESERVES The act included a technical provision clarifying that the invest- ment income on assets held under annuity contracts sold to a gov- ernmental pension plan (whether or not qualified) or to a govern- ment for use under an unfunded deferred compensation plan would be accorded the same tax treatment as income on assets held for qualified pension plans. The effect of the provision was to exclude most of the investment income from Federal income taxation.

4. ROLLOVER OF DISTRIBUTIONS FROM A TAX-SHELTERED ANNUITY ERISA provided that the recipient of a lump-sum distribution from a tax-qualified pension, profit-sharing, stock bonus, or annuity plan could defer tax on the receipt of such distribution by rolling over the proceeds (net of any employee contributions) within 60 days of receipt to an individual retirement account or to another employer-sponsored qualified retirement plan. In lieu of rolling over the distribution, the recipient could elect to have it taxed under a favorable 10-year income averaging procedure. These two options were not made available to persons (employees of tax- exempt organizations and public schools) for whom section 403(b) annuities (so-called tax-sheltered annuities) were purchased or cus- todial accounts were established. This act corrected this "inequity," anomaly, or inconsistency by providing that the recipient of a lump sum distribution from a tax- sheltered annuity could avoid current tax liability by rolling over the distribution within 60 days of receipt to an individual retire- ment account or another tax-sheltered annuity. The amount rolled over to an IRA could subsequently be rolled over again (with accu- mulated earnings) to another annuity contract or custodial account but could not be rolled over into a tax-qualified retirement plan. 5. TRASOP's The Revenue Act of 1978 made the TRASOP provisions a part of the Internal Revenue Code for the first time and stipulated that they would remain in effect until December 31, 1983. It provided that for plan years after December 31, 1978, contributions to a TRASOP would be entitled to a tax credit only if the plan was qualified. It specified that employer contributions had to be allocat- ed in accordance with rules governing tax-qualified plans, stipulat- ing that allocation must be based on employee compensation, taking into account only the first $100,000 of compensation for an employee. Several provisions of the act dealt with the rights of plan partici- pants to vote the shares of the employer stock held under the TRASOP. If a TRASOP holds employer stock of a corporation whose stock is publicly traded, the plan must provide that the stock is to be voted by the plan participants. In addition, the plan must provide for voting and dividend rights equivalent to rights possessed by shareholders of the highest class of stock of the issu- ing corporation "readily available" on a public market. Partici- pants in a TRASOP that hold shares of stock in a closely held cor- poration must be permitted to vote the shares on any issue that by law or charter must be decided by more than a majority vote of common shareholders voting on the issue. A tax qualified defined contribution plan of any type that holds more than 10 percent of its assets in employer securities must permit the participants to ex- ercise the voting rights of the securities held by the plan. For a TRASOP or leveraged ESOP, the only types of employer securities that may be acquired and held by the plan are common stock of the issuing corporation and preferred stock of the issuing corpora- tion that is readily convertible into its common stock. A participant in a TRASOP who is entitled to a distribution under the plan must be given the right to require that the distribu- tion be made in the form of employer securities. The participant may then convert the securities into cash and roll the cash over into an IRA without current tax liability if the rollover takes place within 60 days of the distribution.

6. CASH OR DEFERRED PROFIT-SHARING PLANS By adding sections 401(k) and 402(a)(8) to the Internal Revenue Code, the Revenue Act of 1978 provided that a participant in a qualified cash or deferred arrangement [CODA] would not have to include in income any employer contribution to the plan merely be- cause he could have elected to receive such amount in cash. For a cash or deferred arrangement to be a tax-qualified plan, it would have to satisfy the normal pension plan qualification rules and vest in full all amounts deferred at the participant's election. The act set forth special nondiscrimination rules to test for dis- crimination as to plan participation or contributions to the plan. Under these rules, as clarified in proposed regulations promulgated in 1981 but still not final, a cash or deferred arrangement would meet the nondiscrimination requirements if (1) the actual deferral percentage for the highest paid one-third of eligible employees does not exceed the actual deferral percentage for the other eligible em- ployees by more than 50 percent, or (2) the actual deferral percent- age for the highest paid one-third of eligible employees does not exceed the percentage for the remaining employees by more than three percentage points. If the second test is used, the actual per- centage for the highest paid one-third cannot exceed the actual per- centage for the other employees by more than 150 percent. That is, the higher percentage must not be more than 2.5 times the lower percentage. 7. CAFETERIA PLANS Under a "cafeteria" or flexible benefit plan an employee may chose from a package of employer-provided fringe benefits, some of which may be taxable and some of which may be non-taxable. In this regard, ERISA made a distinction between cafeteria plans in existence on June 27, 1974 and those established after that date. The law stated that an employer contribution made before January 1, 1977 to a cafeteria plan in existence on June 27, 1974 would have to be included in an employee's gross income only to the extent that the employee actually elected taxable benefits. In the case of a plan not in existence on June 27, 1974, an employer contribution would be includible in the employee's gross income to the extent that the employee could have elected taxable benefits. The Tax Reform Act of 1976 extended these rules to contributions made before January 1, 1978. The Revenue Act of 1978 imposed the requirement that a plan be in writing to be subject to cafeteria plan rules and that the employ- ee be given a choice between two or more benefits which could be taxable or nontaxable. To the extent that nontaxable benefits were elected, employer contributions to the plan were not to be includ- ible in the employee's gross income. However, in the case of a highly compensated employee, employer contributions were to be includible in the employee's gross income to the extent that tax- able benefits could have been elected unless the plan met specified nondiscrimination standards as to coverage, eligibility, and contri- butions or benefits. The distinction between plans as to the date of establishment was eliminated. E. MULTIEMPLOYER PENSION PLAN AMENDMENTS ACT OF 1980 5 1. BACKGROUND In the belief that multiemployer pension plans, as defined in- ERISA, rested on a more secure financial foundation than plans sponsored and operated by single employers, Congress imposed a less demanding funding standard on such plans than that applica- ble to single employer plans and set the per capita premium rate for their plan termination insurance coverage at one-half the pre- mium rate for single employer plans. Single employer pension plans established after ERISA were re- quired to fund their supplemental liabilities over a period not longer than 30 years and their actuarial losses over a period not longer than 15 years. In contrast, multiemployer pension plans were permitted to fund their supplemental liabilities over 40 years and their actuarial losses over 20 years. Multiemployer pension plans were required to pay 50 cents per participant per year to the Pension Benefit Guaranty Corporation (hereafter the PBGC) for in- surance protection, whereas single employer plans had to pay $1 per participant (increased to $2.60 per participant in 1978). Multi- employer pension plan premiums and claims payments were to be kept separate from those of single employer plans through the maintenance of two distinct trust (guaranty) funds within the PBGC. The plan termination insurance program established by title IV of ERISA was made effective immediately for a single employer pension plans (certain plan terminations were even covered ret- roactively), but because of the difficulty of determining how the program would function in the multiemployer plan context, Con- gress postponed the effective date of the insurance coverage (but not the obligation to pay premiums to the multiemployer pension plan trust fund) until January 1, 1978 except for discretionary cov- erage under restricted circumstances. In applying plan termination insurance to a multiemployer pen- sion plan, Congress recognized the necessity of distinguishing be- tween termination of the plan itself and termination of plan mem- bership by a signatory employer. No obligation was to attach to the PBGC unless and until the plan terminated, an event thought by the framers of the legislation highly unlikely to occur. Withdrawal of a signatory employer would create no liability for the PBGC but could, depending upon the circumstances, give rise to an obligation to the plan on the part of the withdrawing employer.

2. NEED FOR REMEDIAL LEGISLATION Under the original legislation, an employer that withdrew from a multiemployer pension plan would be liable to the plan for its pro rata share of the plan's unfunded actuarial liabilities for insured benefits, up to 30 percent of its net worth, as defined. The with- drawing employer's share of the unfunded liabilities was to be de- termined by a fraction, the numerator of which was the employer's

5 Public Law 96-364. required contributions to the plan over the preceding 5 plan years and the denominator of which was the aggregate required contribu- tions to the plan by all employers over the preceding 5 years. However, in a major flaw in the legislative design, the employer was obligated to pay this amount to the plan only if the plan itself terminated within 5 years after the employer's withdrawal. A "sub- stantial" employer, defined in the law as one whose required con- tributions during the two immediately preceding plan years, or the second and third preceding plan years, equaled or exceeded 10 per- cent of aggregate required contributions for the same period, was required upon withdrawal from the plan to post a surety bond or escrow assets for the full amount of its allocable share of the plan's unfunded insured liabilities to ensure discharge of its obligation in the event that the plan were to terminate within the next 5 years. But even a substantial employer could escape payment of its con- tingent withdrawal liability if it withdrew from the plan more than 5 years before the plan terminated. In a nutshell, if a multiemploy- er pension plan were to terminate, the current employer members of the plan and those that had terminated within the preceding 5 years would be responsible for the plan's unfunded insured liabil- ities, up to 30 percent of the net worth of the respective employers. In the period between the enactment of ERISA and the sched- uled effective date of multiemployer plan insurance coverage, the basic flaw in the program design became apparent to all concerned parties, as well as the underlying weakness of many multiemployer plans, especially those in declining industries. With the enactment of ERISA, employers signatory to multiemployer plans became con- tingently liable for the first time for the plans' benefit obligations. Prior to ERISA, employers had been obligated only to make their bargained contributions to the plans and had generally been con- tent to leave the plan decisions, including the setting of benefit levels, to the board of trustees, which was traditionally dominated by the labor trustees. Employers generally took the position that contributions to the plan, having been equilibrated to x cents of compensation and paid in lieu of current wages, belonged to the participating employees and should be administered in accordance with the views of the union trustees. Now faced with the possibility of being saddled with their alloca- ble share of the plan's unfunded liabilities, many signatory employ- ers sought opportunities to withdraw from the plan and prospective signatories became reluctant to become a part of the enterprise. The situation was exacerbated by the fact that during the pre- ERISA years many multiemployer plans had funded only the normal cost and interest on the unfunded liability. Even after ERISA, most plans were following a policy of spreading the amorti- zation of the unfunded liability over the full permissible 40-year period. Moreover, prior to ERISA, employers could-and did-with- draw from multiemployer plans and leave their share of unfunded liabilities in the plans, to be borne by the continuing employer members. With the advent of contingent employer liability, pro- spective employer members were being asked to assume a share of the actuarial liabilities left with the plan by long since departed employers, in addition to financing currently accruing benefits. This was not an appealing prospect. It was especially unappealing when the level of employment in the covered craft, trade, or indus- try was in a permanent decline, thus increasing the cost per em- ployee and the burden on the remaining employers. The future of the whole multiemployer plan approach seemed to be in jeopardy. After consultation with all interested parties, including employer associations, labor leaders, congressional staff, administration offi- cials, and pension experts from various backgrounds, the PBGC de- veloped a complex, comprehensive bill that would set aside or modify most of the ERISA title IV provisions pertaining to multi- employer pension plans and substitute a program designed to deal with the unique features and problems of that type of plan. Be- cause of the complexity of the proposed legislation and the difficul- ty of reconciling and accommodating conflicting interests of various concerned groups, it was impossible to get the legislation enacted before the original program was to become effective. Faced with this dilemma, Congress extended the deadline to July 1, 1978. While there was general agreement among all concerned parties on the essential elements of the new approach, there was much contention over some of the critically important details and Con- gress found it necessary to extend the original deadline several times more, the ultimate effective date for the original title IV multiemployer program being set for August 1, 1980. Despite stren- uous efforts on the part of administration and congressional lead- ers to get the amending legislation enacted before August 1, 1980, the deadline was missed and for almost 2 months the original de- fective program was operative, with none of the dire consequences that had been predicted. Finally, on September 26, 1980, after 6 years of study and about 4 years of intense legislative activity by various interest groups, the Multiemployer Pension Plan Amend- ments Act became law.

3. MAJOR FEATURES OF THE ACT

(A) DEFINITION OF MULTIEMPLOYER PLAN ERISA defined a multiemployer pension plan as a collectively bargained plan to which more than one employer contributed but no one employer contributed as much as 50 percent of aggregate contributions in the first year or 75 percent of aggregate contribu- tions in subsequent years. Furthermore, the plan had to assume re- sponsibility for the benefits credited in respect of employers who withdrew from the plan, a form of internal "plan termination" in- surance. The Multiemployer Pension Plan Amendments Act, hereafter re- ferred to as MPPAA, broadened the definition to include all collec- tively bargained plans to which more than one employer contribut- ed, deleting the other two conditions. However, certain plans not previously classified as multiemployer plans but falling within the new definition, were permitted to make a one-time, irrevocable election not to be treated as multiemployer plans.

(B) DEFINITION OF INSURED EVENT The insured event under the original title IV of ERISA was ter- mination of the plan, as defined. Under the amended act, the oper- ational definition for multiemployer plans became plan insolvency or the inability to meet benefit obligations as they become due. Plan termination was no longer to mean dissolution of the plan, with the PBGC assuming responsibility for payment of insured ben- efits. Now termination simply implies freezing of the plan's vested benefits and the cessation of further benefit accruals and further vesting. Employers must continue to contribute to the plan, within limits but without reference to 30 percent of net worth, whether they were still signatory to the plan at time of termination or had previously withdrawn with allocable unfunded liabilities. If the plan, whether formally terminated or not, becomes unable to meet current benefit obligation, the PBGC will advance funds as a loan, which must be repaid if the plan later becomes solvent again.

(C) RETRENCHMENT IN BENEFIT GUARANTEES Title IV of ERISA, as originally enacted, provided that a plan participant's basic benefits, as defined by the PBGC, were to be guaranteed (insured) in full up to $750 per month, with that ceiling to be indexed to the CPI. In order to protect the insurance program against abuse, the law stipulated that insurance protection for newly created benefits, whether arising out of newly established plans or amendments to existing plans, was to be phased in ratably over 5 years; 20 percent of the vested benefits being protected at the end of the first year, 40 percent at the end of the second year, and so on. A minimum of $20 per month was to be phased in each year. It turned out that insurance of $750 per month, indexed, provid- ed protection for 100 percent of the vested benefits of the great bulk of participants in multiemployer plans, which typically credit specified dollar amounts for each year of credited service. Thus, most plan participants would have nothing to lose if their plan were to be terminated by the trustees to relieve the member em- ployers from having to liquidate the plan's unfunded actuarial li- ability, with a similar plan being instituted immediately thereafter to provide future service benefits. The employers would incur li- ability up to 30 percent of their net worth, of course, but many might regard this to be an acceptable or even attractive tradeoff against meeting their allocable share of the plan's unfunded liabil- ities. The losers would be the PBGC and the premium paying em- ployers that did not take advantage of this stratagem. To protect against this potential abuse and to make plan continu- ation attractive to the participants and their union trustees, MPPAA reduced the level of insurance protection per individual to about one-half of that prevailing for participants in single employ- er plans. Specifically, the amended title IV guarantees 100 percent of the first $5 of monthly benefits per year of service but only 75 percent (65 percent for inadequately funded plans) of the benefits in excess of $5, up to $15 per month. In other words, the MPPAA insurance ceiling is $16.25 for each year of service. A particpant with 40 years of credited service could enjoy insurance protection of $650 per month but it would fall short of his total benefit entitle- ment. Benefits in payment status as of July 29, 1980, and the vested benefits of participants within 3 years of normal retirement as of that date were to be guaranteed in accordance with single em- ployer plan rules. As another antiabuse measure, MPPAA stipulated that there would be no insurance protection for benefits in effect for fewer than 5 years. That is, hereafter there is to be no phasing in of pro- tection for benefit improvements in multiemployer plans. The act provided for the creation of a voluntary supplemental in- surance program within 36 months. The purpose of such a program was to restore multiemployer plan insurance protection to a level comparable to that of single employer plans. The law did not spell out the details of the program, leaving that task to the PBGC. Such a program has not been developed as yet.

(D) FUNDING STARDARDS The act contained several features designed to strengthen the funded status of multiemployer plans. First of all, the plans were required to meet the same funding standards for future benefit im- provements as those applicable to single employer plans from the beginning. Supplemental liabilities created by plan liberalizations must be funded over a period no longer than 30 years and actuarial losses must be amortized over a period of 15 years or less. Second, a minimum contribution requirement [MCR] was super- imposed on the basic funding standards. The MCR stipulates that the total contributions to a multiemployer plan (covering the normal costs and amortization of supplemental liabilities) must be sufficient to amortize the unfunded liabilities in respect of retired participants over 10 years and all other unfunded vested liabilities over 25 years. For plans with more retired than active participants, the MCR is reduced by factors reflecting the average guaranteed benefits of retired participants and the ratio of retired to active participants. This reduction is aptly described as the "overburden credit.' A plan entitled to this relief from the minimum funding requirement is simply overburdened with retired participants and a disproportionately high ratio of retired lives liabilities to active lives liabilities. In the same vein, MPPAA made provision for a multiemployer plan to be considered in a state of reorganization. Such a plan must meet certain tests designed to identify a seriously deteriorating fi- nancial condition that could lead to early plan termination if reme- dial measures are not instituted. Reorganization alerts interested parties to potential danger and allows the plan trustees to take action to avoid insolvency. A plan in reorganization may increase contributions or reduce accrued benefits, even those of retired par- ticipants. Benefits can be reduced only if they result from plan amendments adopted after March 26, 1980 and are not guaranteed. This means that such accrued benefits, even when vested, can be reduced down to the level guaranteed by the PBGC, but not beyond.

(E) WITHDRAWAL LIABILITY Perhaps the most important feature of MPPAA-and certainly the most troublesome and controversial-is that which imposes li- ability upon an employer that withdraws from a multiemployer plan. Before the act, a withdrawing employer had only a contin- gent liability to the plan which materialized only if the plan termi- nated within the next 5 years. Under MPPAA, the liability is an absolute one and must be discharged whether or not the plan even- tually terminates. Moreover, the liability is retroactive in that it recognizes benefits that accrued before the enactment of MPPAA or even ERISA. Before ERISA employers had assumed that their obligation to the plan was discharged in full and for all time by payment of the required current contributions. The constitutional- ity of the retroactive feature of the act has been affirmed in a number of court decisions, including one by the U.S. Supreme Court. The recent decision by the Supreme Court 6 affirming the constitutionality of the provision assessing liability against employ- ers that withdrew from multiemployer plans after April 28, 1980- almost 5 months before enactment of MPPAA-was nullified by congressional action.7 As a result of this new legislation, withdraw- al liability will attach only to employers that withdrew after Sep- tember 26, 1980. Congress found it necessary to address four aspects of employer withdrawal liability: (1) The definition of a plan withdrawal, (2) de- termination of the withdrawing employer's allocable share of the plan's unfunded liability for vested benefits, (c) calculation of the withdrawal liability payments, and (4) assessment and collection of the liability payments by the plan. The provisions of the act deal- ing with these four matters are complex and can only be sketched in this chapter.8 (1) Definition of Withdrawal Liability The law distinguishes between a complete withdrawal from a multiemployer plan and a partial withdrawal. A complete with- drawal occurs whenever an employer (1) permanently ceases to have an obligation to contribute to the plan, or (2) permanently ceases all covered operatiois under the plan, including sale of all assets. The first circumstance is generally associated with decertifi- cation of the union that cosponsors the plan, with the employer op- erating thereafter with a nonunion work force or one affiliated with another union. The second circumstance relates to the discon- tinuance by the employer of those operations that were subject to the collective bargaining agreement under which the plan func- tions. Subject to certain exceptions, a partial withdrawal of a signatory employer occurs if any one of the following three conditions is present on the last day of the plan year: (a) The employer's contribution base units (e.g., hours worked) during each of the three consecutive plan years ending with the current. plan year are 30 percent or less of the em- ployer's contribution base units for the high base year, as de- fined. (b) The employer permanently ceases to have an obligation to contribute under one or more, but not all, of the collective

6 PBGC v. R. A. Gray & Co., No. 83-245 (S.Ct., 6-18-84). 'Tax Reform Act of 1984. Public Law 98-369. See Dan M. McGill, op. cit., pp. 565-79 for a fuller treatment of the subject.

37-569 0 - 84 - 5 bargaining agreements under which the employer has been ob- ligated to contribute to the plan but continues to perform work in the jurisdiction of the collective bargaining agreements in question or transfers such work to another location; or (c) The employer permanently ceases to have an obligation to contribute under the plan with respect to work performed at one or more (but fewer than all) of its facilities, but the em- ployer continues to perform work at the facility (or facilities) of the type for which the obligation to contribute ceased. In the building and construction industry (and certain segments of the entertainment industry), there is withdrawal liability only if the employer remains in the area (or returns within 5 years and does not renew its obligation to contribute to the plan) and per- forms work (presumably with nonunion labor) that would have been covered by the plan. There are special rules for plans in the trucking, household goods moving and public warehousing indus- tries. The PBGC is authorized to prescribe regulations under which plans in industries other that construction and entertainment may provide for special withdrawal liabilities rules, if the PBGC deter- mines that the industry characteristics make such rules appropri- ate and that the rules would not pose a significant risk to the PBGC. The term industry characteristics relates primarily to whether or not the amount of work, as measured by contribution base units, is substantially affected by the movement of union em- ployers out of and into the jurisdiction of the plan. Subject to certain conditions, a plan not in the construction in- dustry may provide that an employer can become a member for up to 6 years and then withdraw without any liability for the plan's unfunded vested benefits. This "free look" is intended to help over- come an employer's reluctance to become a member of the plan. There are special rules concerning the sale of assets by a signato- ry employer. These rules are designed primarily for small employ- ers, especially family operated businesses, which frequently sell out to larger firms. Briefly, an employer whose obligation to contribute to the plan is terminated through the sale of its assets in a bona fide, arm's length transaction with an unrelated party is not re- quired to make liability payments if the purchaser of the assets be- comes obligated to contribute to the plan in respect of the oper- ations for substantially the same number of contribution units for which the seller was obligated to contribute and posts a surety bond or escrows assets to assure fulfillment of its obligations. (2) Allocation of Withdrawal Liability The liability of a withdrawing employer may be determined by allocating to it a share of pooled liabilities or by allocating the li- abilities directly attributable to the employer by virtue of the bene- fits credited to its employees during their participation in the plan. The act sets out three acceptable approaches to allocating pooled liabilities, each approach allocating on the basis of plan contribu- tions. A plan may allocate on some basis other than the three stat- utory methods, subject to the approval of the PBGC. Under all pro- cedures liabilities are determined on the basis of vested benefits, whether or not guaranteed by the PBGC. The act specifies the method to be employed, unless the plan adopts another approved method. Under this procedure, called the presumptive method, a distinction is made between the unfunded liabilities for vested benefits for plan years ending before April 29, 1980 and those arising in plan years ending on or after that date. Responsibility for that first pool of liabilities, which includes the unfunded obligations left behind by employers that withdrew over all the years of the plan's existence, is assigned to that group of employers still participating in the plan during the first plan year ending on or after April 29, 1980. Any employer in that group with- drawing before these unfunded liabilities are fully amortized must assume that proportion of the unamortized amount of these liabil- ities which its required contributions over the last 5 plan years ending prior to April 29, 1980, bears to the total plan contributions for that period. For the purpose of this allocation, it is assumed that 5 percent of the original amount of unfunded liabilities are funded each year over a 20-year period beginning with the first plan year ending on or after April 29, 1980. The unfunded vested benefit liabilities arising in plan years ending on or after April 29, 1980 are segregated by-the plan year in which they arise and are prorated among the employers having an obligation to contribute to the plan as of the end of that year. For each plan year ending on or after April 29, 1980, the net change in the unfunded vested benefit liability is computed and prorated over the then participating employers on the basis of contributions for that year and the four preceding plan years. The net change in that liability for a given plan year can arise from (a) experience gains and losses; (b) retroactive plan liberalizations, especially ben- efit increases; (c) reallocation of withdrawal liabilities of earlier plan years that could not be collected or were not assessed because of various relief provisions; and (d) assumed amortization of the prior year's liability. As with the pre-April 29, 1980 unfunded liability pool, each plan year's unfunded liability is assumed to be reduced by 5 percent of the original amount per year through contributions. Thus, each pool has a life of 20 years. This produces a "rolling" liability attri- bution with a "piece" of the total potential liability of an employer expiring after the lapse of 20 years. Stated differently, this method provides for 21 distinct liability pools: the pre-1980 pool and one for each of 20 plan years. As would be surmised, the other two statutory pooled liability al- location approaches provide for different pooling arrangements. The modified presumptive method provides for only two pools, one for pre-April 29, 1980 unfunded liabilities and for liabilities arising in plan years ending on or after that date. The second pool is allo- cated among the participating employers on the basis of contribu- tions to the plan over the preceding 5 years, irrespective of when the employers became signatories to the plan. The third method of allocating pooled liabilities, the rolling five approach, makes no distinction between the pre-April 29, 1980 li- abilities and those arising thereafter. All unfunded liabilities, irre- spective of the plan year or years of origin, are pooled and allocat- ed on the basis of contributions over the preceding 5 years. The un- funded liability is recomputed each year, taking into account past amortization payments and withdrawal liability payments. There is no assumed rate of amortization, the computation reflecting actual experience. As an alternative to pooling unfunded liabilities and allocating them on a basis that achieves only rough equity, a plan may elect, if feasible, to identify the source of the actuarial liabilities and at- tribute them (with matching assets) to the employers in respect to whom they arose. In concept, this "direct attribution" method is the most equitable procedure for allocating liabilities, but it is fea- sible only for plans in industries enjoying reasonably stable em- ployment relationships. Even under this method, a withdrawing employer must assume its allocable share of the pooled unattribu- table liabilities of previously withdrawn employers. (3) Withdrawal Liability Payments MPPAA embodied the general concept that a withdrawing em- ployer should discharge its withdrawal liability, not by a lump-sum payment, but through a series of periodic payments at the approxi- mate level at which the firm was contributing to the plan during the years immediately preceding withdrawal. The amount of the installment payment is determined by reference to the contribution base and the contribution rate. The contribution base is unique to the withdrawing employer, being the average annual number of contribution base units for the three consecutive plan years for which the employer's contribu- tions were the highest during a period of 10 consecutive plan years ending before the plan year of withdrawal. For most plans, a con- tribution base unit is an hour of covered employment. This rather harsh definition of the contribution base was intended to discour- age employers from deliberately winding down operations before withdrawal to minimize their liability payments. It may work a hardship on an employer suffering a genuine, unavoidable decline in operations during the last few years before withdrawal, a not unlikely occurrence. The contribution rate of a multiemployer plan is negotiated by the collective bargaining representatives, the same rate generally applying to all employees signatory to the plan. The rate to be used in calculating an employer's withdrawal liability payment is the highest rate at which the employer was obligated to contribute during the 10-year period ending with the plan year in which the withdrawal occurs. The plan year of withdrawal was included in the measurement period since, under recent experience, the high- est contribution rate is likely to be found in that year. The employer's annual withdrawal liability payment is arrived at by multiplying the contribution rate times the contribution base. The employer is obligated to make a payment of this amount each year until its withdrawal liability is fully amortized but, in no event, for a period of more than 20 years, even though the liability is not fully amortized. In the event that a withdrawn employer be- comes insolvent before discharging all of its withdrawal liability, the plan has a claim against the bankrupt estate for the unamor- tized portion of the liability. Under the de minimis rule, enacted for the relief of small em- ployers, especially those wishing to sell their assets or go out of business, an allocable withdrawal liability of less than $50,000 is waived and somewhat larger amounts can be partially waived. An employer's obligation for a partial withdrawal is determined by multiplying the withdrawal liability for a complete withdrawal by the ratio of contribution base years in the year following with- drawal to the average contribution base units during the 5 years preceding the withdrawals. (4)Assessment and Collection of Withdrawal Liability Under MPPAA, it is the responsibility of the plan trustees to as- certain the amount of the withdrawing employer's liability, inform the employer of the amount of its liability, and demand payment. Determination of the withdrawal liability is a complex matter, in- volving, inter alia, an actuarial valuation of the plan's vested liabil- ities and assets, and must be carried out by the plan's enrolled ac- tuary under the direction of the plan trustees. The actuary must use actuarial assumptions that in the aggregate are reasonable and represent his best estimate of future experience, and he must use actuarial techniques that are appropriate to the task. The employer has 90 days after receipt of the notice of assess- ment in which to contest the amount of the liability or the sched- ule of payments. However, within 60 days the employer must make its first quarterly payment, even though the amount or legitimacy of the payment is in dispute. If the employer questions the amount of the assessment, the schedule of installment payments, or the le- gitimacy of the assessment, and the differences cannot be resolved by the two parties, the disputed matters must go to arbitration. For the purpose of the arbitration proceeding, the plan's determi- nation of the amount of withdrawal liability and the schedule of payments is presumed to be correct unless the employer can show 'by a preponderance of evidence that the determination was unrea- sonable or clearly erroneous." If the dispute involves actuarial con- siderations, to have liability assessment itself set aside or modified the employer must prove that the actuarial assumptions and meth- ods employed in the determination of the liability were unreason- able or that the plan's actuary made a significant error in applying the actuarial assumptions or methods. The arbitrator's decision may be appealed to the Federal courts by either party within 30 days. Withdrawal liability payments are made quarterly, unless the plan specifies a different pattern. The employer may repay the out- standing amount of unpaid liability at any time. Interest is charge- able for late payment and default in payment of a quarterly in- stallment can result in the outstanding amount of liability becom- ing payable in a single sum, a penalty known as "acceleration." 4. TERMINATION INSURANCE PREMIUMS MPPAA increased the per capita premium for multiemployer pension plans from 50 cents to $1, effective immediately, and pro- vided for graded annual increases that would bring the rate up to $2.60 by 1989, the rate in effect for single employer plans at the time MPPAA was approved. The act stipulated that if the projected rate increase proves to be inadequate, the PBGC could propose re- ducing benefit guarantees instead of increasing premiums.

5. SINGLE EMPLOYER PLANS MPPAA made several changes that affected single employer plans, as well as multiemployer plans. The contingent employer li- ability insurance [CELI] provisions of ERISA, designed to provide insurance to employers against the assessment of liability for un- funded insured benefits, were repealed in their entirety, upon rec- ommendation of the PBGC staff and its Presidentially appointed advisory committee. The act authorized the Secretary of Labor to issue regulations al- lowing cost-of-living increases for retired persons to be treated as a welfare plan rather than a feature of a qualified pension plan. The principal significance of this action was to relieve cost-of-living ad- justments from minimum funding requirements, maximum benefit limitations, and plan termination insurance protection and associ- ated employer liability. F. ECONOMIC RECOVERY TAX ACT OF 1981 [ERTA] 9

1. KEOGH PLANS AND SEP's ERTA boosted the deduction limit for employer contributions to a defined contribution Keogh plan, to a defined contribution plan maintained by a subchapter S corporation, and to a simplified em- ployee pension [SEP] from $7,500 per year to $15,000. For all three types of plans, the 15 percent of compensation limit was retained. To provide a similar increase in the benefit limit for a defined ben- efit Keogh or subchapter S corporation plan, the compensation taken into account in determining permissible annual benefit ac- cruals was raised from $50,000 to $100,000. The act increased from $100,000 to $200,000 the annual compen- sation that may be taken into account in determining maximum permissible annual contributions to defined contribution Keogh, subchapter S, and SEP plans. However, if the plan recognizes com- pensation in excess of $100,000 per year, the contribution rate for common-law employees must not be less than 7/2 percent of com- pensation. For subchapter S corporation plans, the act extended to all part- ners the existing Keogh plan rule that the use of an interest in the plan as security for a loan is to be treated as a distribution. Finally, the act authorized a penalty-free correction of an excess contribution to a Keogh plan if the excess is withdrawn before the filing date of the tax return.

2. INDIVIDUAL RETIREMENT ACCOUNTS ERTA provided a significant impetus to the individual retire- ment plan approach to old age economic security by broadening the

9Public Law 97-35. eligibility provisions to permit employees and self-employed indi- viduals already covered by other retirement programs to establish IRA's. Under the original enabling legislation, the privilege of es- tablishing individual retirement plans was confined to persons not already participating in an employer sponsored qualified plan or other tax-favored arrangement. ERTA also raised the annual con- tribution limit from the lesser of $1,500 or 15 percent of compensa- tion to the lesser of $2,000 or 100 percent of compensation. The limit for a spousal IRA was increased from $1,750 to $2,250 and the existing requirement that contributions under a spousal IRA be equally divided between the spouses was deleted, subject to the pro- vision that no more than $2,000 be allocated to the account of either spouse. As a result of these liberalizing changes and more aggressive marketing of their investment services by financial in- stitutions, about $32 billion was set aside in IRA's in 1983. Under existing law, individuals could generally direct their own investments in the sense that they could select among a number of investment vehicles or under certain arrangements even direct the acquisition of a particular item of property. This sometimes led to abuses, especially when individuals invested their IRA accounts in "collectibles" (antiques, art, gems, stamps, etc.) and used the ob- jects while they were being held as investments. ERTA dealt with this problem by providing that any amounts in an IRA account in- vested in collectibles, as defined, would be treated as distributions for income tax purposes. The act stipulated that the proceeds of a redeemed U.S. retire- ment bond distributed under a qualified bond purchase program could be rolled over, tax free, to an individual retirement account. U.S. retirement bonds purchased for an employee could be re- deemed only after the individual attained age 59 , died, or became totally disabled. The act allowed a deduction to a divorced spouse for contribu- tions to a spousal IRA established by the individual's former spouse at least 5 years before the divorce if the former spouse contributed to the IRA under the spousal rules for at least 3 of the 5 years pre- ceding the divorce. If these conditions are met, the limit on the di- vorced spouse's IRA contributions for a year is not less than the lesser of $1,125 or the sum of the divorced spouse's compensation and alimony includible in gross income.

3. QUALIFIED VOLUNTARY EMPLOYEE CONTRIBUTIONS ERTA also made provision for the deduction of voluntary em- ployee contributions to a pension plan, profit-sharing plan, or a savings program classified as "qualified voluntary employee contri- butions" [QVEC's]. These deductible contributions may be made to a qualified pension plan, a section 403(b) annuity, a SEP-IRA, or to a Government plan, whether qualified or not. The maximum amount that can be deducted each year is $2,000, the same as for an IRA. Active participants in qualified plans may split their con- tributions between plans and IRA's as long as the total amount contributed does not exceed the specified limit. QVEC's are permit- ted only if the particular plan allows them. If a plan allows QVEC's, it must allow the participant to designate a contribution as not being a QVEC. A distribution of accumulated deductible con- tributions from a qualified plan is taxed as ordinary income, the same as a distribution from an IRA. Tax deductions were not allowed for mandatory employee contri- butions. These were defined as contributions required either as a condition of employment, a condition of participation in the plan, or a condition of obtaining additional employer-provided benefits. Employee contributions to a savings plan that are matched at some percentage by the employer have been construed to be mandatory contributions and hence not eligible for a tax deduction. However, voluntary, unmatched contributions to a savings plan are eligible for the $2,000 deduction. 4. EMPLOYEE STOCK OWNERSHIP PLANS [PAYSOP's] ERTA brought about a major shift in the tax treatment of ESOP's by substituting a payroll-based tax credit for the invest- ment-based tax credit that had been available since 1975. The in- vestment-based credit served as an incentive to the creation and maintenance of a TRASOP for only those firms that tend to be cap- ital intensive and make substantial investments in plant and equipment. With the intent of providing incentives to a broader segment of industry, including labor intensive companies, congres- sional ESOP advocates persuaded their fellow legislators to provide a tax credit based on payroll, the resulting arrangement generally known as a PAYSOP. The Tax Revenue Act of 1978 had made pro- vision for the investment credit to run through 1983 but ERTA had it expire one year earlier. The new law set the maximum tax credit at one-half of 1 percent of covered payroll for 1983 and 1984 and three-quarters of 1 per- cent of covered payroll for the years 1985 through 1987. The tax credit was to expire on December 31, 1987. This credit was not to be available if contributions allocated to officers, shareholders, and highly compensated employees (those whose earnings are more than two times the dollar limit on employer contributions to de- fined contribution plans) exceeded certain limits. ERTA increased from 15 percent of covered payroll to 25 percent the deduction allowed the employer for contributions to an ESOP where the contributions are to be applied to make principal pay- ments on a loan incurred to purchase employer stock. An unlimit- ed deduction was made available for contributions applied to pay interest on the loan. The act also specified that (1) contributions applied to pay loan interest, and (2) forfeitures of fully leveraged ESOP stock were to be disregarded for purposes of existing limitations on employer con- tributions that may be added to an employee's individual account in any 1 year (at that time, the lesser of $41,500 or 25 percent of compensation). This rule was to be applicable only if the contribu- tions allocated to the proscribed group of participants does not exceed certain limits. The act permitted distributions from a tax-credit ESOP made on account of the sale of corporate assets or the distribution of a sub- sidiary to be made in less than the 84-month period otherwise ap- plicable to distributions of employer stock. Finally, the act deleted with respect to securities acquired after 1979 for profit-sharing plans the existing requirement that an em- ployee be permitted to vote stock allocated to his account. The re- quirement was continued for all other defined contribution plans. G. TAX EQUITY AND FISCAL RESPONSIBILITY ACT OF 1982 [TEFRA] 10 Enacted with relatively little fanfare and public debate, the pen- sion-related provisions of TEFRA were of enormous significance and have had a material impact on all types of pension programs. The most significant modifications in existing law included: -Sharp reductions in maximum allowable limits under qualified defined benefit and defined contribution plans. -Onerous new requirements for topheavy plans. -Liberalizations in Keogh plan requirements. -Mandatory tax withholding from retirement benefits unless the recipient opts out. -Restrictions on the time period over which survivor benefits, other than to a spouse, can be paid. -Limitation on the amount of pension (or IRA) death benefits, whether payable as an annuity or lump sum, that may be ex- cluded from Federal estate taxation. -Restrictions on loans to participants. -Change in Social Security integration rules for defined contri- bution plans.

1. MAXIMUM LIMITS Prior to ERISA, there were no statutory limits on benefits or contributions on qualified pension plans. On the theory that there should be some limit on the use of "tax expenditures' to provide employer-sponsored pensions, ERISA introduced a benefit limit of the lesser of $75,000 per year or 100 percent of the individual's av- erage compensation during the preceding 3 years. For a defined contribution plan, that "reform' legislation imposed an annual contribution limit of the lesser of 25 percent of compensation, or $25,000. Both dollar linmits were indexed to the CPI. By 1982, the defined benefit plan limit had risen through indexation to $136,425 and the defined contribution plan limit to $45,475. TEFRA reduced the defined benefit plan dollar limit to $90,000, effective in 1983, and the defined contribution plan dollar limit to $30,000. The respective percentage-of-compensation limits remained unchanged. The dollar limits were to remain frozen for 1984 and 1985, but beginning in 1986 they were again to be linked to the CPI. The 1986 adjustment was to be based on the ratio of the index for the fourth quarter of 1985 to the index for the fourth quarter of 1984. In addition to the foregoing reductions in dollar benefit and con- tribution limits, TEFRA reduced the combined limits when an em- ployer sponsors both a defined benefit and a defined contribution pension plan. The existing combined limit was 1.4 times the allow- able limits. TEFRA reduced that multiple to a limit of 1.25. For

10Public Law 97-248. topheavy plans (see below), the 1.25 limit was reduced to 1 unless certain conditions are met. TEFRA provided transitional rules which had the practical effect of not requiring any adjustments in benefits accrued under the pre- vious limits nor in contributions already made. For these individ- uals, no additional benefits were to accrue until the maximum limit, as adjusted for cost-of-living-increases, exceeds the partici- pant's current accrued benefit. ERISA permitted the maximum dollar benefit to be paid under any annuity form, including a 100 percent joint and survivor annu- ity, beginning as early as age 55. TEFRA specified that beginning in 1983, the dollar limit under defined benefit plans was to be actu- arially reduced for payments starting before age 55. The interest rate used to determine the actuarial reduction cannot be less than the higher of 5 percent or the rate specified in the plan. The higher the interest assumption the greater the reduction in benefits. The act stipulated that the required adjustment for benefits com- mencing before age 62 was not to have the effect of reducing the maximum dollar benefit below $75,000 for benefit payments com- mencing between ages 55 and 62. For payments beginning before age 55, the dollar limitation is to be reduced to the equivalent of a $75,000 benefit at age 55. This means, of course, that the prescribed actuarial reduction for early retirement will have no practical effect until 1986 or later when indexation is expected to raise the maximum dollar benefit above $75,000. If the benefit payments start after age 65, the plan may provide for an increase in the dollar limitation-in addition to any cost-of- living increases. The interest rate used to determine this upward adjustment cannot exceed the lower of 5 percent or the rate speci- fied in the plan. In this case, the lower the interest assumption the lower the boost in benefits. The dollar limitation must be adjusted actuarially if the benefit is to be paid out under any annuity form other than a straight life annuity or a qualified joint and survivor annuity. The interest rate used in this calculation cannot be less than the higher of 5 percent or the rate specified in the plan.

2. TOPHEAVY PLANS TEFRA established some special, restrictive rules for what it termed "topheavy" plans, which can be either defined benefit or defined contribution plans. These rules apply equally to plans of corporations, partnerships, and proprietorships, the distinctions among such plans having been largely eliminated. Under TEFRA, a defined benefit plan is considered to be top- heavy if more than 60 percent of the actuarial present value of the accrued benefits of all employees and their beneficiaries under the plan is attributable to "key employees" or their beneficiaries. Key employees include officers (up to 50); employees owning at least 1 percent of the firm and enjoying annual compensation of more than $150,000; 5 percent owners irrespective of compensation; and the 10 employees owning the largest interest in the employer. Any person meeting any of those tests at any time in a plan year is treated as a key employee. A defined contribution plan is considered topheavy if the sum of the account balances of participants who are key employees ex- ceeds 60 percent of the account balances of all employees under the plan. For the purpose of making these determinations, accrued bene- fits and account balances include distributions made within the preceding 4 years. If a plan is topheavy, the maximum compensa- tion that can be taken into account for setting benefits and contri- butions is $200,000, indexed beginning in 1986. A topheavy plan must provide at least minimum benefits or con- tributions for all participants who are not key employees. For years in which a defined benefit plan is topheavy, nonkey employ- ees must accrue a minimum nonintegrated benefit of at least 2 per- cent of final average pay (5 highest consecutive years) per year for up to 10 years. Similarly, for years in which a defined contribution plan is topheavy, the employer contribution for participants other than key employees must generally equal at least 3 percent of pay. Topheavy plans are subject to the TEFRA imposed limits on ben- efits or contributions, including the combined dollar limits. Howev- er, the combined plan dollar limit is 1 (not 1.25) unless (1) the de- fined benefit provides a minimum benefit accrual of 3 percent per year up to 10 years and the defined contribution plan calls for an employer contribution of at least 4 percent of compensation, and (2) key employees' accrued benefits or account balances do not account for more than 90 percent of the actuarial value of accrued benefits or the aggregate account balances of all employees under the plan. In addition, topheavy plans must provide for accelerated vest- ing-either full vesting after 3 years or graded vesting providing 20 percent at the end of 2 years of service and grading up in annual increments of 20 percent until full vesting is applicable after 6 years. Distributions to a key employee under a topheavy plan must begin in the tax year in which the key employee reaches age 70/2, even if the individual continues to work for the employer. Also, if any benefits are distributed to a key employee under a topheavy plan before age 591/2, other than by reason of death or disability, there is a nondeductible penalty tax of 10 percent of the amount of the distribution. 3. KEOGH PLANS One of the purposes of TEFRA was to achieve parity between corporate section 401(a) plans and Keogh plans for the self-em- ployed. To this end, most of the special rules applicable to Keogh Plans were removed for tax years beginning after 1983. Maximum limits were increased effective in 1984 in line with the new $90,000/$30,000 limits for corporate plans. This represented at least a twofold increase in permissible benefits and contributions under Keogh plans. These plans were made subject to the new top- heavy rules but were relieved of some special requirements that were even more onerous. As a result of the removal of the special restrictions on Keogh plans: -Benefits no longer have to be vested immediately. -The limitations on benefits to owner employees no longer apply. -Past service benefits can be provided under a defined benefit Keogh plan on the same basis as for a corporate plan. -Coverage and participation rules can be the same as for corpo- rate plans. -Social Security integration rules are the same as for corporate plans. -Voluntary employee contributions up to 10 percent of compen- sation are permitted even when only owner-employees partici- pate in the plan. -The 6 percent excise tax on excess contributions no longer ap- plies. -There is no longer a 5-year restriction on reentry into the plan for an employee who receives a premature distribution. -Plan assets do not have to be held by a bank or financial insti- tution. -The first $5,000 of lump-sum death benefit can be excluded from Federal income taxation.

4. TAX WITHHOLDING As part of an overall effort to achieve greater tax compliance and enhance the cash-flow of the Federal Government, TEFRA re- quires that income tax be withheld from most taxable pension and annuity benefits in excess of $450 per month, unless the recipient elects not to have taxes withheld. Lump-sum distributions, in-serv- ice withdrawals, and deferred compensation payments are subject to the same rules. All recipients of income benefits must be provided annually with notice of the right to make, renew, or revoke an election not to have taxes withheld. If a recipient elects to have no withholding, that election remains in effect until the recipient revokes it. In this connection, TEFRA includes a provision requiring plans to maintain records sufficient to determine the tax treatment of distributions and to report relevant information to participants and the IRS. For any year the plan does not maintain adequate records, the plan administrator is subject to a penalty of $50 for each indi- vidual for whom the administrator fails to maintain records, up to a maximum of $50,000. No penalty is to be imposed for the inad- equacy of pre-1983 records where a reasonable effort is made to cor- rect the inadequacy. 5. DEATH BENEFITS TEFRA stipulated that effective with plan years beginning after December 31, 1983, payment of death benefits to a survivor other than a spouse will have to be completed within 5 years after the participant's death. The only exception to this rule is where bene- fits are payable under a term certain option and the term does not exceed the life expectancy of the employee or the joint life expect- ancy of the employee and spouse. Benefits to a spouse can still be spread over her remaining life. This restriction applies to all qualified plans and IRA's. 6. ESTATE TAX EXCLUSION Prior to TEFRA, employer-provided death benefits payable under a qualified plan to a beneficiary other than the decedent's estate where fully excludible from Federal estate taxation. TEFRA im- posed a limit of $100,000 on such exclusion for both qualified plans and IRA's. The adverse impact of this action was mitigated by the availability of the marital deduction if the beneficiary is a surviv- ing spouse.

7. LOANS Under TEFRA, loans from tax-qualified or governmental defined benefit or defined contribution plans (including tax-sheltered annu- ities) are to be treated as plan distributions for Federal income tax purposes, unless certain amount and repayment requirements are met. Loans in amounts of less than $10,000 and those less than $50,000 that do not exceed one-half of the present value of the par- ticipant's vested benefits are not to be treated as distributions if the terms of the loan agreement call for repayment in less than 5 years. A loan not required to be repaid within 5 years will be treat- ed as a taxable distribution irrespective of amount. An exception to the 5-year repayment rule was made for a loan used to acquire, construct, or substantially rehabilitate the princi- pal residence of the participant or a member of the participants' family. In that event, a "reasonable" repayment schedule in excess of 5 years is allowed.

8. SOCIAL SECURITY INTEGRATION Congress considered a major revamping of the statutory and reg- ulatory rules pertaining to the integration of defined benefit and defined contribution plans but ended up changing only the simple, straightforward rule for integrating defined contribution plans. Prior to ERTA the differential between the employer's contribution rate with respect to earnings above and below the Social Security wage base could be as high as 7 percent. ERTA reduced the differ- ential to the employer's payroll tax rate for OASDI, excluding med- icare, consistent with the presumption underlying integration of defined benefit plans that the employer contribution toward an em- ployee's Social Security income benefits "purchases" half of those benefits. That tax rate is 5.4 percent of the Social Security wage base in 1984, rising to 5.7 percent for the years 1985-89. The rate is scheduled to increase to 6.2 percent in 1990 and remain there in- definitely.

9. OTHER PROVISIONS

(A) FUNDING TEFRA confirmed that anticipated future cost-of-living adjust- ments to the overall dollar benefit limits could not be taken into account in determing the allowable deduction for employer contri- butions to defined benefit plans. This had been the position of the IRS. (B) DISABLED PARTICIPANTS IN DEFINED CONTRIBUTION PLANS The act authorized employers to continue contributions to de- fined contribution plans for employees (other than officers, share holders, and highly compensated individuals) who become perma- nently and totally disabled. The contributions must be based upon the employee's annualized compensation at the onset of disability and be fully and immediately vested. This provision removed a dis- parity between defined contribution and defined benefit plans as to the permissible treatment of disabled participants, it having been the sanctioned practice of defined benefit plans to continue the ac- crual of benefits for disabled participants.

(C) SEP'S Simplified employee pensions were made more attractive (and more nearly comparable to conventional corporate defined contri- bution plans) by increasing the limits on deductible contributions to the lower of 15 percent of compensation or $30,000 (indexed). Thus, over an 8-year span, the maximum pemissible deduction for an employer contribution to an employee's SEP account went from $1,500 per year to $30,000. The requirement for a minimum contri- bution of 7 /2 percent of compensation when earnings above $100,000 are recognized, was rescinded.

(D) SUBCHAPTER S CORPORATIONS Shareholder employees of subchapter S corporations are to be treated the same as any other employees, but, as would be expect- ed, the plans will be subject to the new topheavy rules. Sharehold- er employees will no longer be taxed on excess employer contribu- tions over the current Keogh limitations on their behalf and they will be eligible to participate in forfeiture allocations under defined contribution plans. For all intents and purposes, these plans are hereafter to be accorded the same treatment as conventional corpo- rate plans.

(E) ROLLOVER OF IRA'S Prior to this legislation, the entire distribution from an IRA had to be rolled over to be tax-free. TEFRA authorized the tax-free roll- over of any portion of an IRA distribution. On the negative side, the act stipulated that beneficiaries (other than the surviving spouse) of deceased individuals will not be permitted to roll over IRA's that they inherit in taxable years beginning after 1983. H. TAX REFORM ACT OF 1984 " This is a massive, complex tax measure designed to increase Fed- eral revenues, primarily through the closing of tax loopholes. It contains more than 40 provisions, many of a liberalizing nature, re- lating to employee benefits, including welfare plans, pension plans, and employee stock ownership plans. This review is limited to pen- sion plans and ESOP's.

I Public Law 98-369. 1. PENSION PLAN PROVISIONS

(A) TIMING OF DISTRIBUTIONS For tax-minimization and general estate planning purposes, many pension plan participants would like the option of delaying the commencement of benefit payments, or distribution of an ac- count balance, for an indefinite period-even to death and beyond. Recognizing this manifestation of adverse selection against the tax system, Congress has long insisted that plan distributions com- mence not later than a specified, determinable date, a heavy non- deductible penalty excise tax otherwise being assessed. The general rule has been that a plan distribution must be made (or install- ment benefit payments commence) by the end of the plan year in which the participant reaches age 70 or retires. The distribution could be made over the remaining lifetime of the participant (or the remaining joint lives of the participant and spouse) or over a period certain no longer than the life expectancy of the participant (or the joint life expectancy of the participant and spouse). The new act made two important changes in the rules relating to required distributions. Under previous law, installment distribu- tions could not extend beyond the life expectancy of the participant (or joint life expectancy of the couple) as calculated on the date the distribution began. The new law permits the single or joint life ex- pectancy to be recomputed each year, thus producing a "rolling" life expectancy. This has the effect of stretching out the payments and ensuring that the participant (and beneficiary) does not outlive his income. The other significant change permits the participant to choose someone other than the spouse as the joint annuitant. Now the participant may choose a child or sibling, which not only provides more flexibility in estate planning but could spread the payments over a longer period. As a safeguard against the participant select- ing a real young person as joint annuitant, the law stipulates that at least half of the account balance must be .distributed over the participant's life expectancy. Those two changes were made applicable to distributions upon the participant's death. The general rule still is that a plan, as a condition for qualification, must require that any remainder inter- est be distributed within 5 years of the participant's death. The rule was modified in two ways by the recent act. The death distri- bution must begin within 1 year of the participant's death, or at such later date as IRS regulations may specify. If the beneficiary is the spouse of the deceased participant the distribution need not commence until April 1 following the year the decedent would have reached age 701/2, unless the surviving spouse dies in the in- terim, in which case the 1-year-of-death rule applies. All of the foregoing changes apply not only to qualified employ- er-sponsored plans but also to IRA's, Keogh plans, and section 403(b) annuities. At the other end of the distribution spectrum, the new law tight- ened the restrictions on premature distributions to certain plan participants. Under prior law, distributions to a key employee in a topheavy plan before he reached age 591/2, became disabled, or died were subject to a nondeductible 10 percent excise tax penalty. Under the new law, the penalty tax is assessed against premature distributions to a 5 percent owner, irrespective of whether the plan is top heavy. As before, the penalty tax is applied only to prema- ture distributions attributable to the years the participant was a 5 percent owner (or, previously, a key employee).

(B) TAX-FREE ROLLOVER OF PARTIAL DISTRIBUTIONS It has been the practice of some small employers to pay retiring employees their pension or profit-sharing accumulation in several installments over a period longer than a year. Such installments have not been eligible for a tax-free rollover into an IRA or 10-year forward averaging. The new law permits a tax-free rollover to an IRA if the distribution in a single tax year represents at least 50 percent of the participant's interest in the plan. For purposes of ap- plying the 50 percent test, amounts credited to the participant under other qualified plans or section 403(b) annuities of the same employer are not aggregated. The participant must elect to have the tax-free rollover treatment apply since there are some poten- tial disadvantages associated with it. The money involved cannot be rerolled over to another qualified plan, the remaining balance loses its eligibility for long-term capital gain treatment (of pre-1974 accumulations) and 10-year forward averaging, and any net unreal- ized appreciation on distributed employer securities, even if attrib- utable to employee contributions, is taxed in the year of distribu- tion.

(C) TOPHEAVY PLANS The new law contains a number of changes to the section 416 topheavy rules. A technical correction to TEFRA was made by sub- stituting employee for participant in the general definition of a key employee. It excluded from the definition of key employee any offi- cer whose annual compensation is no more than 150 percent of the limit on annual additions under defined contribution plans. Cur- rently, that means a salary of no more than $45,000. The individ- ual involved may be classified as a key employee under one of the other tests but not because he is an officer. For purposes of determining the 10 employees owning the largest interests in the employer firm (one of the four alternative tests of a key employee status), the law provided that only employees with annual pay in excess of the dollar limit on annual additions to de- fined contribution plans is to be taken into account. If two employ- ees have the same ownership interest, the one with the greater annual pay is to be treated as having the larger interest. Under the so-called 5-year lookback rule, aggregate distributions made to any employee within the 5-year period ending with the de- termination date must be counted when measuring an employee's account balance or the actuarial present value of the employee's cumulative accrued benefits for topheavy purposes. The new law clarified that distributions from a terminated plan must also be taken into account under the 5-year lookback rule. The new law also stipulated that for plan years beginning after 1984 the accrued benefits of any individual who has received no compensation (other than benefits) directly from the employer for a 5-year period ending on the determination date are to be ignored in testing for topheavy status. As another form of restriction on key employees in topheavy plans, the new law does not permit tax-free rollovers to an IRA and thence to a new employer plan if any part of the distribution is attributable to contributions made while the individual was a key employee in a topheavy plan. There seems to be no prohibition against a direct transfer of the employee's inter- est from the trustee of the old employer's plan to the trustee of the new employer's plan. Another amendment exempted governmental plans from the top- heavy qualification requirements on the premise that they do not carry the same abuse potential as private sector plans. Finally, the new law requires IRS to publish by the end of 1984 final regula- tions on topheavy plans or "boilerplate" topheavy plan language that plans can use as a "safe harbor" for up to 6 months after the final regulations are published.

(D) SEP'S Several provisions of the new law made changes in the rules gov- erning SEP's. The pre-TEFRA definition of earned income for SEP's (and Keogh plans) was restored by providing that earned income was to be computed without taking into account deductible employer contributions. Though TEFRA raised the dollar limit on employer contributions to a SEP from $15,000 to $30,000, it did not increase the corollary $15,000 limit on employee deductions for em- ployer contributions (which are treated as income to the employee). This oversight was corrected by the new law. The employee deduct- ible limit is now $30,000. As a requirement for qualification, TEFRA generally requires that topheavy defined contribution plans make minimum contribu- tions on behalf of nonkey employees of 3 percent of pay. The new law extended this requirement to SEP's. Finally, the new law provides that if the employer does not maintain any other integrated plan during the taxable year, em- ployer OASDI payroll taxes may be treated as employer contribu- tions to a SEP if the practice is uniformly applied to all employees.

(E) REPEAL OF ESTATE TAX EXCLUSION The act repealed the $100,000 estate tax exclusion for distribu- tions from qualified corporate plans, SEP's, IRA's and Keogh plans in respect of persons dying after December 31, 1984. An unlimited sum can continue to be bequeathed to a spouse without tax liability through the marital deduction. Thus, it is worthy of note that pay- ments to minor children of the decedent shall be treated as paid to a surviving spouse if those amounts become payable to her upon the child's death or attainment of majority.

(F) ALIMONY AND IRA'S Except for spouses of employed persons, an IRA can be estab- lished and maintained only by an individual with earned income. Contributions do not have to be made from earned income but they

37-569 0 - 84 - 6 are limited to 100 percent of such income. In order to make it pos- sible for nonworking divorced spouses (assumed to be predominant- ly women) to make contributions to an IRA, the new law treats all taxable alimony as compensation for IRA purposes. (G) WITHDRAWAL LIABILITY UNDER MULTIEMPLOYER PENSION PLANS As stated earlier, MPPAA was enacted September 26, 1980, but its withdrawal liability provisions applied retroactively to with- drawals after April 28, 1980. The constitutionality of the retroac- tive feature was tested in the Federal courts and was ultimately upheld by the Supreme Court, just days before enactment of the Tax Reform Act of 1984. In recognition of this holding, the law eliminated retroactive liability and provided for refunds of any amounts paid, less reasonable administrative expenses incurred in refunding the payments. Despite these refunds, the law stipulates that the withdrawal liabilities of other employers are not to be in- creased to make up the loss to the plans. (H) PLAN LOANS TO PARTICIPANTS The new law clarifies the rules governing plan loans to partici- pants after August 13, 1982 by making it clear that any participant can borrow up to $10,000, even if that sum exceeds one-half of his vested interest in the plan, without the loan being treated as a tax- able distribution. Loans from deductible employee contributions are treated as taxable distributions, regardless of their amount. (I) PLAN CONTRIBUTION AND BENEFIT LIMITS As was noted earlier, TEFRA required certain actuarial adjust- ments to the $90,000 limit for benefits commencing after age 65 or before age 62. Apparently, there was confusion in the minds of some as to whether the required adjustments were to the benefits themselves or merely the dollar limit. The new law made it clear that the required adjustments are to the dollar limit on annual benefits. (J) DEADLINE FOR IRA CONTRIBUTIONS Under the new law, a contribution to an IRA must be made not later than April 15. Under previous law, the contribution could be made within the period covered by an extension (or extensions) of the individual's tax return filing date. (K) EXTENSION OF FREEZE ON BENEFIT AND CONTRIBUTION LIMITS TEFRA reduced the overall dollar limits on contributions and benefits under qualified plans and supended cost-of-living adjust- ments to those limits until 1986. The new law suspends adjust- ments to the limits until 1988. 2. CASH-OR DEFERRED ARRANGEMENTS [CODA's] Prior to this latest legislation, a cash or deferred arrangement, also known as a CODA or section 401(k) plan, was subject to special tests under which the amount that highly compensated employees could elect or defer (and exclude from current gross income) de- pended upon the level of deferrals by other employees, the compar- ison being in terms of the percentage of compensation deferred. These special tests did not permit employer contributions to Social Security to be taken into account. The proposed regulations issued in 1981, however, permitted a section 401(k) plan to satisfy the gen- eral nondiscrimination standards for qualified plans in lieu of the special tests. These general standards permitted employer Social Security contributions to be taken into account. The new legislation resolved this conflict by stating that a CODA is not a qualified CODA unless the special tests applicable to such an arrangement are satisfied. If an arrangement fails to satisfy the special tests, amounts deferred are not excluded from gross income, but the plan of which the arrangement is a part is not be be dis- qualified if it meets the usual rules for qualification, including the general tests for nondiscrimination. The deferred percentage taken into account for purposes of applying the special tests is the sum of the deferral percentages for an employee under each CODA in which the employee participates. In other words, the plans are to be aggregated for the purposes of applying these tests. The conference report endorsed the rule in the proposed regula- tion concerning recognition of nonelective contributions of the em- ployer to a CODA in determining whether the special deferral per- centage tests have been met. Such contributions may be taken into account only if they (1) are nonforfeitable when made, and (2) satis- fy the withdrawal restrictions pertaining to elective deferrals under a qualified CODA. 3. EMPLOYEE STOCK OWNERSHIP PLANS In the new law, Congress provided some new inducements to the formation of ESOP's. To encourage banks, insurance companies, and corporations actively engaged in the business of lending money to make leveraged ESOP loans, the law provides a 50 percent ex- clusion for interest received or accrued on loans to leveraged ESOP's to acquire employer securities. The loan may be made di- rectly to the leveraged ESOP or to a sponsoring corporation that in turn relends the proceeds. Lenders are expected to pass on some of the tax savings to the ESOP's through lower interest charges. It has been estimated that lenders could cut their interest rate by 2 or 3 percentage points. This makes an ESOP an attractive vehicle for financing a leveraged buyout of another company, probably an unintended result of the legislation. A group planning a takeover could create an ESOP, ostensibly for the benefit of the employees of the target concern, and have the ESOP acquire some or most of the shares of the target company through subsidized loans. The planning group could exercise control over the target company with a relatively small percentage of the outstanding voting stock that it holds directly, since the employee participants can not vote the shares of stock held on their behalf by the ESOP trustee. The new law also allows a special corporate deduction for cash dividends paid with respect to stock held on the dividend record date by a PAYSOP or an ESOP, provided the dividends are paid directly in cash to participants or indirectly to them through the plan, within 90 days after the close of the plan year. These divi- dends are fully and currently taxable to the individual plan partici- pants. Under previous law, corporations did not realize gain or loss on the contribution of employer securities to a PAYSOP to the extent that the contribution was a precondition to a tax credit. The new law eliminates that treatment of PAYSOP's but extends it to ESOP's provided that after the "sale" the ESOP owns at least 30 percent of the total value of the employer securities and within a 15-month period (beginning 3 months prior to the transfer) the em- ployer purchases "qualified replacement securities." The corpora- tion's gain or loss potential is preserved through an adjustment to its basis in the qualified replacement securities. ERTA set the maximum tax credit for employer contributions to a PAYSOP at one-half of 1 percent for years 1983 and 1984 and three-quarters of 1 percent for years 1985 through 1987. The new legislation freezes the credit at one-half of 1 percent through 1987, at which time this special tax incentive will expire unless extended by Congress in the interim. Finally, the new law imposes a 10 percent excise penalty on cer- tain premature distributions of employer securities from a PAYSOP or ESOP within three years of receipt. 4. CAFETERIA PLANS Flexibility in the employee benefit program of an employer has always been a desirable characteristic, enabling the participating employees to adapt the available set of benefits to their own special needs and circumstances. It has become an even more sought-after attribute in recent years as the traditional role of women in family and economic life has undergone a transformation, if not revolu- tion, and other fundamental changes have occurred in the social and political fabric of the country. The most obvious approach to providing the desired flexibility is to permit plan participants to pick and choose among a number of employer-sponsored benefit plans, the election being effective for a meaningful period of time. If deemed essential by the employer, all employees could be required to take a basic core of benefits, the election applying only to supplemental layers of benefits or to types of benefits not regarded to be essential to all employees. These arrangement came to be known as "cafeteria" plans. The major concern over these plans, apart from their additional admin- istrative complexity, was whether under the doctrine of construc- tive receipt a participant in a nondiscriminatory cafeteria plan would be treated as having received a taxable benefit solely be- cause he had the opportunity, before the benefits became available to him, of choosing between the taxable and nontaxable benefits under the plan. This concern was laid to rest by ERTA in an amendment to section 402(a)(1) of the Internal Revenue Code. As a logical extension of the cafeteria concept, many employers set up an arrangement for certain types of expenses of a current nature under which an individual account is maintained for each eligible employee, and at the beginning of each plan year the em- ployer credits each account with a stipulated or agreed upon sum of money, with any covered expenses incurred during the year being reimbursed from and charged against the account. These ar- rangement are usually referred to as "flexible spending accounts" or "benefit bank" flexible spending accounts. Some of these ar- rangements provide that any unused account balances will be paid in cash to the employees as additional taxable wages. Other such arrangements make no provision for payment of the unused bal- ance. These are usually termed "use it or lose it" plans, and they are believed by some to encourage overutilization of benefits to avoid the loss of the unused balance. The flexible spending account type of cafeteria plan provided the springboard for an even more innovative-and controversial-ar- rangement, called zero balance reimbursement accounts or ZEBRA's. Under this approach the employer does not credit the employee's account in advance or even specify the level of benefits that will be provided under the various elected benefits. Instead, the employee is simply reimbursed by the employer as the covered expenses are incurred, the amounts expended being deducted from cash wages. In this manner, the employee's taxable compensation is reduced by the amount of any covered expenses incurred and no tax is paid on the reimbursement checks from the employer (in the absence of a tax ruling or law to the contrary). After some intimations of unhappiness over certain of these ar- rangements, the Internal Revenue Service on February 10, 1984 issued a news release (IR 84-22) which stated in effect that an ar- rangement which assures a participant that he will receive his full compensation, whether as a consequence of expense reimburse- ment, cash refund, or absence of covered claims, does not provide nontaxable benefits under the code. In simple terms, under such an arrangement any expenses reimbursed to a participant by the em- ployer or any amount credited irrevocably to his account, would have to be included in the participant's current taxable income. In May 1984, IRS issued proposed regulations intended to imple- ment and enforce the stand taken in the news release. The regula- tions identified the type of benefits that could be provided under a properly constituted cafeteria plan on a nontaxable basis and those that could not be provided under such a plan. They confirmed the message of the news release by stating that benefits under a cafete- ria plan would not be excludible from the participant's gross income if the participant were assured of receiving the equivalent of the benefits whether or not he incurred covered expenses. The regulations, if they were to become final, would have undesirable tax consequences for cafeteria plans that provide for cash outs of account balances; carryover of contributions or benefits from one plan year to the next; application of credits to purchase benefits for later plan years; revocation of benefit elections during a given "benefit election period"; and benefit election periods so short or frequent that an employee can time his elections to cover predict- able expenses or claims. ZEBRA's and most benefit bank flexible spending accounts would effectively be banned under the proposed regulations. The Tax Reform Act of 1984 affirmed the proposed regulations but stipulated that the final version should not be applied retroac- tivity to any cafeteria plan in existence on February 10, 1984, the date of the IRS news release alerting the business community to the questionable features of some cafeteria plans. The proposed rules will generally become effective January 1, 1985, except for plans that are amended before that date in which event the earlier date applies. Plans have until July 1, 1985 to eliminate a money back feature from an existing plan. Under the new legislation, only benefits that are specifically ex- cluded from gross income under the Internal Revenue Code can be provided under a cafeteria plan. Group term life insurance, health care benefits, group legal services, dependent care, and possibly dis- ability income benefits are permissible benefits. Certain nontaxable benefits, including scholarships, vacation facilities, physical fitness clubs, van pooling, summer camps, and parking, cannot be included in a cafeteria plan. The plan cannot include any benefit that defers compensation except for the opportunity afforded participants to make elective contributions under a CODA. Key employees participating in a cafeteria plan will be taxed as if they had received all available taxable benefits under the plan if more than 25 percent of the total nontaxable benefits are provided to key employees. This determination is to be made on the basis of the value of the coverage rather than actual expense reimburse- ment or loss payment. Several influential members of Congress feel that the legislation should not have banned the "money back" feature of flexible spending account plans and are cooperating with certain segments of the business community in efforts to have that feature restored. I. CONCLUDING OBSERVATIONS It can be seen from this review of pension legislation since ERISA that an extensive body of law has been enacted during the last decade, greatly expanding many of the regulatory concepts of that historic piece of legislation. The melange of provisions present- ed in seriatim fashion may have been confusing to some who are not intimately familiar with the various pension arrangements and the basic framework of law within which they operate. Yet several broad congressional concerns or policy objectives are discernible in the succession of laws enacted. These policy objectives appear to be: -To standardize the tax and other treatment of the qualified pension arrangements, such as conventional corporate plans, Keogh plans, SEP-IRA's, and section 403(b) annuities. -To encourage individual retirement arrangements to narrow the uncovered segment of the active labor force. -To promote equity among the covered groups of employees through various strictures designed to minimize discrimination in favor of owner-employees, officers, and highly compensated individuals. -To establish realistic limits on benefits that can be provided or contributions that can be made under tax-favored arrange- ments. -To preserve the multiemployer pension plan mechanism, while retaining a meaningful benefits insurance program for that important type of pension arrangement. -To encourage employee ownerhip of the employing firm through employee stock ownership plans. At the time of this writing, Congress had just completed consid- eration of the Retirement Equity Act of 1984 which is primarily de- signed to provide more equitable pension treatment of women. Among other things, the act will lower to 21 the age at which pen- sion benefits must begin to accrue after 1 year of service; provide survivor annuity protection before the participant is eligible for early retirement; exempt court-ordered payments to a divorced or separated spouse from the ERISA prohibition against assignment or alienation of pension benefits; and to liberalize break-in-service rules to recognize women's special childrearing responsibilities. For several years legislation has been before Congress that would make important changes in the single employer plan termination insurance program, including substituting employer insolvency for plan termination as the insured event. The legislation would also increase the premium from $2.60 per participant per year to as much as $8.50. At one time passage of this legislation seemed certain-even im- minent. In recent months the support of both organized labor and the business community has weakened, for different reasons. The ERISA Industry Committee, which had been the leading business proponent of termination insurance reform over the past several years, has withdrawn its support from existing House and Senate bills. Thus, prospects for early enactment of this much-needed leg- islation appear dim. Attention is now being turned to a separate bill or amendment to another bill to increase the premium rate. Also, for the last several years there has been some interest in Congress in regulating (to a limited extent) pension plans of States and local government units. Under the proposals to date, federally- imposed standards would be confined to reporting and accounting, avoiding the more sensitive substantive areas of discrimination, vesting, funding, investments, and the like. Notwithstanding, all such proposals have, in the main, been opposed by State and local government lobbying groups. The latest bill is called the Public Employee Pension Plan Reporting and Accountability Act [PEPPRA]. It is currently on the legislative backburner and seems likely to remain there for some time. Chapter 4 IMPLEMENTATION OF ERISA IN THE EXECUTIVE BRANCH (Prepared by Beverly M. Klimkowsky,I Ph.D., Public Policy Consultant Specializing in Income Maintenance Policy) A. INTRODUCTION President Ford's signing of the Employee Retirement Income Se- curity Act [ERISA] into law on September 2, 1974 marks both an ending and a beginning. Responsibility for pension reform shifted from the legislative branch of government to the executive branch. Members of Congress, pleased to have been able to reach a consen- sus after many years of toil and acrimony, had completed their task and congratulated themselves for taking historic steps to safe- guard pensions. The job was just beginning for the executive agen- cies charged with administering the new law. ERISA imposed upon them the routine tasks faced when implementing any piece of regu- latory legislation. However, its complexity, inconsistencies and cumbersome administrative apparatus added such additional bur- dens that it is a minor miracle that ERISA was ever implemented at all. This chapter focuses on the implementation of ERISA in the ex- ecutive branch. Thus, the official departure point is the signing of the bill into law. However, the roots of implementation begin during the policy formation process when political decisions deter- mine the essential parameters of what is to be implemented, who is to implement it, and when it must be done. 2 By necessity, some ref- erence will be made to events preceding ERISA's enactment. It should be noted at the outset that describing ERISA's imple- mentation is a subject whose scope easily could span a book rather than this single chapter. Therefore, this author has condensed much material gleaned through interviews with members of the pension community and secondary analysis of congressional hear- ings, oversight reports and a review of the pension literature.

, Dr. Klimkowsky has worked as a policy analyst for the Social Security Administration and the Maryland State Office on Aging. She received her doctoral degree from the Johns Hopkins University in 1982 after completing a dissertation entitled "Assessing ERISA Amid Changing Expectations." The author wishes to acknowledge the following persons who permitted her to interview them in 1982 and 1984 and gave unstintingly of their time: Paul Berger, Jeffrey Clay- ton, Gerald Cole, Karen Ferguson, Michael Gordon, Donald Grubbs, Jeffrey Hart, Travis Knight, Ian Lanoff, Alvin Lurie, Edward Lysczek, Walt Marciniak, Jerry Oppenheimer, Steven Sacher, Lawrence Smedley, Dallas Salisbury, Leslie Shapiro, Ira Shepard, and Peter Turza. Several other persons also granted interviews, but preferred to remain anonymous. The author is in- debted to them as well. 2 For a more detailed discussion of the events which preceded ERISA's enactment, see chap- ters 1 and 2. Writing about ERISA's implementation in a simple, systematic manner poses problems because of the overlapping jurisdictional apparatus Congress designed. Three agencies share primary re- sponsibility for implementing ERISA. One could treat each agency separately, essentially writing three papers; or take a chronological approach, showing what is happening simultaneously in each agency. Although problems exist with each approach, a chronologi- cal treatment has been chosen for this analysis since implementa- tion can be broken into several stages and the agencies charged with ERISA's implementation interact with each other. Before describing ERISA's implementation, it is important to un- derstand what is meant by program implementation. Therefore, a brief overview of implementation, the stages which comprise it, and the conditions which enhance it will be made so ERISA's im- plementation can be compared with "normal" implementation for other programs. B. PRINCIPLES OF IMPLEMENTATION Implementation is the process by which a decision is carried out. In the public sector, the initial decision usually is a law which an agency is charged with carrying out. Implementation includes es- tablishing the agency to carry out the task, translating the broad goals into operational goals, and doing what is necessary to achieve those goals, whether it is building a tunnel or issuing forms which are to be returned. In the case of ERISA, the primary goal is to insure that persons receive the private pensions which have been promised to them under conditions which are not unduly stringent. Administering a law is often thought of as an automatic process accomplished by faceless bureaucrats after Congress has deter- mined a course of action. Schoolchildren learn that the legislative branch "makes laws" and that the executive branch "administers the laws." This dichotomy between "politics" and "administration" formed the base for the early literature in public administration 3 and persists today despite many studies showing the inaccuracies and inadequacies of it.4 Implementation has been a topic of study for about fifteen years. Policy analysts and scholars of public administration have identi- fied stages of implementation and conditions necessary for success- ful program administration. The stages of implementation will form the framework for this analysis and the conditions for imple- mentation will be used as criteria to evaluate ERISA's implementa- tion in the executive branch. Program implementation has several distinct phases: an initial startup period; a "settled dust" phase of full implementation; and a more mature, policy-oriented stage. The tasks needed to be done, the degree of organization, and the expectations people have for these different phases distinguish them.

3 Goodnow, Frank. Politics and Administration. New York, The Macmillan Co., 1900; Wilson, Woodrow. The Study of Public Administration. Political Science Quarterly, vol. 56, December 1941. pp. 486-506. 4 Simon, Herbert A. Administrative Behavior: A Study of Decision-making Processes in Ad- ministrative Organization. New York, Macmillan, 2d ed. 1961; Dahl, Robert A. The Science of Public Administration: Three Problems. Public Administration Review, vol. 7,1947. pp. 1-11. During the startup phase of implementation, two major goals must be accomplished: an administrative apparatus must be estab- lished and the standard operating procedures necessary to adminis- ter the law must be instituted. The former condition is a necessary, but not sufficient predecessor of the latter. Accomplishment of both of the above goals marks the beginning of the settled dust stage of full implementation. In a regulatory program, most of the major rules and regulations have been pro- mulgated and the agency processes paperwork routinely. Although some gaps undoubtedly will exist, they are relatively minor or affect a small portion of the universe covered. Some programs never reach the mature, policy-oriented stage of implementation. Taking an active role in policymaking and being recognized as an expert characterize this phase of implementation. Tracking criticism for a program can be used to indicate what phase of implementation the program is in. In the startup phase, criticism centers on the program itself and in the delay in getting the program operational-delays in issuing rules, regulations, in- terpretive bulletins, and forms. In the settled-dust phase of normal implementation, complaints focus more on onerous paperwork re- quirements, service delays, technical inconsistencies of the program or enforcement delays. During the mature phase of implementa- tion, concern about individual case problems continue, but more global concerns emerge. For example, criticism about the adminis- trative apparatus or the leadership of an agency might emerge. In all stages, legislators and officials must satisfy several condi- tions before the law can be carried out successfully. Edwards and Sharkansky note: They must issue policy directives that are clear and con- sistent; hire adequate staff and provide them with the in- formation and authority necessary to carry out their orders; offer incentives for staff to execute policy as deci- sionmakers intended; and effectively followup on the im- plemented actions of subordinates.5 Ideally, all of the conditions are met. Realistically, most pro- grams suffer deficiencies in one or more areas. The severity of the shortfall determines how many problems implementors will en- counter. Failure to implement a program occurs on a continuum. Fortunately, implementors usually can overcome problems if the desire and will to administer the program exist. Congress plays an important part in all of the above steps. It issues policy directives in the laws that it writes, makes provision for scarce resources, and grants authority to carry out the delegat- ed duties. Congress also leads oversight activities to insure that its programs are being implemented as it desires. Before examining the stages of ERISA's implementation in light of the above criteria, an overview will summarize the stages ERISA implementation has passed through over time.

5 Edwards, George C. III and Ira Sharkansky. The Policy Predicament and Implementing Public Policy. San Francisco, W.H. Freeman and Co., 1978. p. 142. C. OVERVIEW: IMPLEMENTATION OF ERISA In terms of the stages outlined above, one would categorize the implementation of the Employee Retirement Income Security Act of 1974 [ERISA] as being in the second of the three stages-the set- tled-dust stage. Pressure currently builds for a move to be made to a more mature phase of implementation. When President Ford signed ERISA into law on Labor Day 1974, the administrative apparatus charged with implementing the new law reflected the ambiguity concerning the proper jurisdictional sphere for the law and continued congressional rivalry over turf. Pensions bridge both tax and labor law. Since the revenue acts of the 1920's pensions traditionally have fallen under the aegis of the Internal Revenue Service [IRS] because of their tax-qualified status. However, since they also represent deferred compensation to workers, they are an important labor issue as well. Congressional concern about pensions began when the effects of abusive pension practices manifested themselves in the denial of pensions to workers with long service records and in the lack of adequate funding of many plans. The Senate Labor Committee under the leadership of Senators Jacob Javits and Harrison Wil- liams began to investigate the issue and uncovered widespread abuses. As time passed and pension reform reached the top of the legisla- tive agenda, tax and labor committees in both the House and Senate held hearings and sponsored pension reform bills. Tax com- mittees wanted the IRS to administer ERISA; labor committees supported the Department of Labor [DOL]. Neither constituency would cede control over pensions by deferring to the other or by establishing an independent agency to administer ERISA. Intransigence characterized the conference which met to recon- cile the differences in three pension bills. Neither House nor Senate, Labor nor Tax, wanted to yield to another. The different parties "compromised" by putting all three bills into different titles of ERISA. Not surprisingly, the result of this "compromise" is a very lengthy bill filled with internal inconsistencies. Agreeing to shared jurisdiction for ERISA administration repre- sented another major congressional compromise. Most experts be- lieve that ERISA never would have passed had it not been for the dual jurisdiction compromise. A former aide to the Senate Labor Committee spoke of his nocturnal meeting with his counterpart of the Tax Committee to convince him to accept shared jurisdiction. 6 Although the Senate conferees preferred divided jurisdiction, they ceded to the House Committee on Ways and Means which refused to consider anything but overlapping jurisdiction. Primary responsibility for administering the law was divided among the ERISA triplets-the Internal Revenue Service, the De- partment of Labor, and the newly formed Pension Benefit Guaran- ty Corporation [PBGC]. Smaller responsibilities were charged to the Departments of Commerce, Health and Human Services, and Jus- tice; the Equal Employment Opportunity Commission; the Securi- ties and Exchange Commission; the Small Business Administration;

6Interview with Michael Gordon. the Federal Trade Commission; and the National Labor Relations Board. The IRS and DOL shared responsibility for implementing many of the same provisions of ERISA. The coordination required to agree upon every regulation and exemption bogged down imple- mentation so much that it proved to be unworkable. The original administrative apparatus Congress established in ERISA was jetti- soned in 1978 by President Carter's Reorganization Plan No. 4. The startup phase for ERISA in IRS and DOL lasted from 1974 through 1978. Once the reorganization plan divided responsibilities between the agencies, obviating the need to coordinate policy, each agency was able to move quickly into the settled dust phase of full implementation where they remain today. The timing for the PBGC differs because of the novelty of its mandate. Although the agency successfully organized an adminis- trative apparatus in a short period of time, it ran into serious prob- lems implementing several provisions of its mandate because major policy decisions needed to be made. Uncertainty shrouded the agency until 1980 when major legislative changes were enacted. Thus, the startup period in the PBGC lasted from 1974 through 1980. Once major program difficulties had been resolved, the PBGC stabilized and currently is in the settled dust of normal implemen- tation. Notably, ERISA implementation has not reached the mature stage of implementation in which the administering agencies act as powerful players in the policy process. To their credit, the agencies charged with implementing ERISA have successfully untangled a complex law and simplified an unwieldy administrative apparatus. Most persons would agree that the law is being administered fairly smoothly, although a few problems remain in enforcement, the ex- emption process, and unissued regulations. However, none of the agencies administering ERISA has consistently assumed a leader- ship position in pension policy which most recently has been influ- enced by the tax committees on the hill, women's groups, and the investment community-none of which views the fulfillment of the promise for a private pension as its sole or primary concern. D. STAGE 1: STARTUP PHASE

1. GETTING ADMINISTRATIVE APPARATUS IN PLACE Fulfilling the pension promise is the raison d'etre of ERISA. However, the idealized verbal goal must be translated into subgoals which a nuts and bolts bureaucracy can administer. Establishing an organization to administer the new law is the first step. Although implementation officially begins when a bill is signed into law, advanced planning for implementation can occur prior to that time. Such planning can ease the problems encountered in the startup phase since the time to implement the law has been effec- tively lengthened. The duration of the startup phase varies depend- ing upon the complexity of the law, time available for transition, and the internal organization necessary to gear up the administra- tive apparatus. ERISA specified different guidelines for the administrative appa- ratus of the three major implementing agencies. In title I, ERISA charged the Labor Department with numerous responsibilities, but did not indicate how administration would be structured. In title II, ERISA charged the IRS with its duties, but created a new office, Employee Plans/Exempt Organizations [EP/EO], under the direc- tion of a new Assistant Commissioner. In the third case, title IV established the PBGC as an independent corporation within DOL, administered on a daily basis by an Executive Director, but nomi- nally headed by a board of directors chaired by the Secretary of Labor. Thus, two distinct types of bureaucratic organizations ad- minister ERISA: two structures placed within preexisting depart- ments and one entirely new structure in the form of a semiautono- mous corporation. The tasks ERISA assigned to DOL and the IRS were much great- er and more complex than those assigned the PBGC. Both bodies faced the problem of determining how to integrate the new pro- gram into their existing structures and of coordinating policies with each other.

(A) INTERNAL REVENUE SERVICE As is often the case, the Congress consulted the Internal Revenue Service and the Department of Labor while ERISA was being writ- ten to get their input on how each wanted to administer the pro- gram. The IRS approached ERISA's implementation with its usual professionalism. Since pensions diverge from the main function of the IRS, i.e., revenue raising, competition did not arise within the Service to determine who would administer ERISA. It determined early that the best way to carry out its obligations under the new law would be to establish a new component headed by a new As- sistant Commissioner. Before ERISA was signed into law, Alvin Lurie, who had been selected as the first Assistant Commissioner for Employee Plans/ Exempt Organizations, came to Washington and served as a con- sultant who paved the way for implementation. Personnel who had been working with pensions were transfered into the new office which had three major components: Employee Plans, Exempt Orga- nizations, and Actuarial. Once ERISA passed, IRS got its supergrade officials in position quickly and was ready to deal with the substantive parts of the program within a year. However, despite the advantages of ad- vanced planning, trained personnel, and extensive support from other offices at IRS and Treasury, EP/EO faced a formidable task. Time to organize, interpret the law, promulgate regulations, design forms, and train field personnel was short. Under preferred circumstances, IRS's national office formulates rules and regulations which it then transmits to its regional and district offices; it also trains field personnel so that the law can be administered in a uniform manner. However, when ERISA was passed, the national office and the field offices recognized that the field offices would be responsible for implementing ERISA before policy directives could come down from Washington. Thus, the na- tional office developed a strategy which emphasized publishing in- terpretive and implementing materials and placed a high priority on promulgating the regulations which would be needed first.7 The IRS also relied upon outside information sources like the BNA Pen- sion Reporter to interpret ERISA.8

(B) DEPARTMENT OF LABOR Faced with many of the same responsibilities as IRS, DOL ap- proached ERISA's implementation much less systematically. Inter- nal bureaucratic rivalries hindered implementation from the initial planning stages. When Congress asked DOL what type of adminis- trative apparatus it desired, Labor officials squabbled beyond the point at which a structure could be stipulated in the law. Since the Labor Management Services Administration [LMSA] had adminis- tered the Welfare and Pension Plan Disclosure Act which ERISA supersedes, it wished to administer ERISA despite its primary re- sponsibility to administer the Landrum-Griffin Act. Other factions within DOL wants ERISA to be administered in a new office under a new Assistant Secretary, at a level on par with that at IRS. The tumult continued for 3 years under five ERISA administrators. In November 1974, the Secretary of Labor delegated authority to administer ERISA to the Assistant Secretary of Labor-Management Relations. He established the Office of Employee Benefit Security within LMSA, charged with overall program administration and enforcement. In May 1976, the Secretary of Labor divided the ERISA program within this Department. LMSA retained field and management op- erations and systems services, but program administration and policy development was spun off to the Office of Pension Benefit Welfare Programs [PWBP], formerly the Office of Employee Bene- fit Security. Not surprisingly, having two offices within DOL administering ERISA resulted in "a lack of clear lines of authority, disagreement over the allocation of personnel and resources and bickering over (various) matters." 9 In September 1977, the Secretary of Labor issued Order 9-77 which reintegrated PWBP into LMSA. The As- sistant Secretary for Labor Management Relations once again had direct authority for ERISA and the PWBP Administrator reported to him.1 0 Three years after President Ford signed ERISA into law, DOL finally settled upon an administrative structure that would last for 7 years. Needless to say, much of the time which Labor officials should have spent hiring qualified staff members, designing forms, pro- mulgating regulations, and processing exemptions was spent argu- ing about who was to administer ERISA in what setting. Within 3 years, five persons assumed the position of the Administrator of ERISA. Implementation strategies and priorities changed as fre-

7 U.S. Congress. House. Committee on Education and Labor. Subcommittee on Labor Stand- ards. Oversight on the Employee Retirement Income Security Act of 1974. Hearings, 94th Cong., 1st Sess., April 29-May 6, 1975. Washington, U.S. Govt. Print. Off., 1975. p. 470. a Interview with IRS official. * Testimony of Francis X. Burkhardt. U.S. Congress. Senate. Committee on Labor and Human Resources. Subcommittee on Labor. Oversight of ERISA. Hearing, 95th Cong., 1st Sess., April 11, 13, and 18, 1977. Washington, U.S. Govt. Print. Off., 1977. p. 69. 1oThe current PWBP administrator has succeeded in having PWBP removed from LMSA as of May 1984, changing the administrative arrangement once again. quently. As backlogs mounted, many sources criticized DOL for its internal disorganization and inability to carry out its duties in a timely manner. 1'

(C) PENSION BENEFIT GUARANTY CORPORATION The PBGC surmounted its initial organizational difficulties more gracefully than Labor. During the summer of 1974, a planning group with representatives from Labor, Commerce, Treasury, and the Office of Management and Budget met to lay the groundwork for the new corporation.1 2 Within 1 week of ERISA's enactment, the PBGC Board of Directors had met, adopted initial bylaws, and borrowed $100,000 from the Treasury; 36 people (most of whom where detailed from other agencies) were at work in newly rented offices. Since premiums were due for defined benefit plans within 30 days from enactment, the PBGC had to mail out over 100,000 pre- mium forms immediately. Although staff members wrote the forms and mailed them out within 30 days, they extended the due date for initial premium receipt by 2 months. Developing operating procedures to process notices of intent to terminate plans was the second major item on the PBGC agenda. During the first 16 months, the agency received almost 6,000 termi- nation notices, vastly exceeding initial expectations. Many of the plans had assets sufficient to pay their liabilities, so the agency fo- cused on processing those terminations first.

(D) SUMMARY ERISA's three main agencies accomplished their internal organi- zation with varying degrees of success. B'oth the IRS and PBGC en- gaged in planning prior to ERISA's enactment, an activity which seemed to give them an edge in gearing up quickly. The IRS relied on well-organized national and field offices to provide backup sup- port. At the PBGC, a small cadre of competent people, aided by the PBGC Advisory Committee, worked long hours to put the program in place. Because of the lack of preexisting rivalries, a sense of teamplay revailed. DOL was unable to plan effectively because of internal disagreements within the agency. This internal turmoil exacerbated the problems it faced during the startup phase. While good organization does not preclude implementation diffi- culties, it does prevent scarce resources from being siphoned into nonproductive activities. Regardless of the level of success attained in organizing, all three of ERISA's implementing agencies experi- enced serious implementation problems during the initial phase.

2. OVERVIEW OF PROBLEMS ENCOUNTERED DURING THE STARTUP PHASE To assess the severity of the problems ERISA administrators faced, it would be useful to return to the criteria for evaluation laid out previously. Whether or not these conditions have been met becomes very apparent in the startup phase of implementation.

I U.S. General Accounting Office. Efforts to Implement the Employee Retirement Income Se- curity Act of 1974 by the Department of Labor; report to the Senate Committee on Human Re- sources by the Comptroller General of the United States. Washington, 1977. 20 p. 12 Testimony of Steven Schanes. Senate Committee on Labor and Human Resources. Over- sight of ERISA. p. 630. 88 As a first condition, Congress ideally states its policy directives clearly and unambiguously. More often, policy directives neither specify what Congress intends nor conform with each other. Com- promises which are needed for a bill to be passed often permit con- flicting passages to be written into law. Administrators of the law must interpret ambiguity and iron out conflicts. As one of the most complex laws Congress ever passed, ERISA suffers from having an unclear mandate. Multiple jurisdiction is a major example of congressional indecision being papered over and left to the administrators to sort out. Some of ERISA's provisions (e.g., paperwork) are too specific, leaving administrators with little flexibility. Many other provisions were so vague that over 100 regu- lations needed to be issued. The second condition which enhances successful implementation is provision of resources adequate to do the job. In addition to per- sonnel and authority, these scarce resources include time and in- formation. The staff must be adequate in both size and training. The authority delegated to an administering agency must enable it to carry out its duties unencumbered. Congress again plays a cru- cial role in supplying adequate resources. It must appropriate the needed funds before employees can be hired. Congress often stints on the provision of sufficient time to carry out the charged tasks. Legislators often expect laws which have taken years to enact, to be implemented in a very short period of time. Agencies usually procure information on their own; however, Congress plays a vital part by appropriating the funds necessary for research. In the case of ERISA, IRS had an advantage in obtaining skilled personnel since it already employed most of the Government's pen- sion experts and only had to transfer them to a new unit. DOL, on the other hand, faced huge program responsibilities without a base core of expertise upon which to rely. Its workers were more famil- iar with law enforcement and processing forms than with pension plans. The PBGC did not have any experts in pensions, but re- quired a smaller number than either IRS or Labor. Time was in short supply when ERISA passed into law. As is often the case, Congress spent over 10 years formulating its policy, but expected its directives to be turned into reality post-haste. By examining table 1, one can see that the effective and due dates of many of ERISA's provisions are short, especially when considering the complexity of the legislation.

TABLE 1.-EFFECTIVE AND DUE DATES OF ERISA PROVISIONS

Provision Effectiveduedate'

Reporting anddisclosure...... January 1, 1975. Participation andvesting ...... Decem ber31, 1975. Funding...... Decem ber 3 1, 19 75. Fiduciary responsibility ...... January 1,1 975. Individual Retirement Accounts ...... Decem ber 31, 1974. Single-Employer term ination insurance...... June 30, 1974. Multiem ployer term ination insurance ...... January 1, 1978. Joint study onpublic retirem entplans...... December 31, 1976. Establish joint boardfor enrollment of actuaries...... October31, 1974. 1Effective date for most plans in existenceonor before January 1, 1974. Source:Employee Retirement Income Security Act. 89 Overlapping jurisdiction between IRS and DOL violated the con- dition that staffs be given the authority necessary to carry out their orders. Lodging a program within a preexisting department further compromises a staff's autonomy because it must complete for resources and the ear of the Secretary. Studies of implementa- tion have shown that programs which have a high number of deci- sion points fail to be implemented successfully more often than programs which require fewer layers of review before a decision is finalized.1 3 ERISA's administrative apparatus certainly required that a high number of decision points be passed through-a precon- dition indicating potential implementation problems. A staff with the desire to implement a program is another pre- requisite for successful implementation. Congress can offer incen- tives and sanctions to insure that administrators carry out its in- tentions. Stable leadership in an agency enhances successful pro- gram implementation since scarce resources need not be spent edu- cating new personnel and reorganizing the agency to suit the new leaders. In the case of ERISA, it was difficult to offer incentives to the staff for policy to be executed as decisionmakers intended because of the lack of clearly stated intentions in the law and legislative history. ERISA's complexity defied substantive comprehension by all but a few legislators. Most legislators who supported ERISA knew that they wanted the pension promise to be kept without unduly restrictive standards. However, they did not have the ex- pertise to evaluate what the standards should be, much less specify how the standards should be operationalized. Attracting qualified leadership proved to be easier for the IRS than for the PBGC or DOL. To date, only two persons have held the job of top administrator of ERISA at IRS, while seven have served at Labor and eight at the PBGC.

TABLE 2.-TENURE OF OFFICIALS RESPONSIBLE FOR ADMINISTERING ERISA

Position Official Tenure

PWBPAdministrator (DOL)...... Monks...... 1984 to present. Clayton ...... 1981-1983. Lanoff ...... 1977-1981. Ballard...... 1977. Chadwick...... 1976-1977. Hutchinson...... 1975-1976. Ballard...... 1974-1975. Assistant Commissioner,EP/EO (IRS) ...... Winborne...... 1979 to present. Lurie...... 1974-1979. Executive Director, PBGC...... Tharp...... 1 984 to present. Jones...... 1982-1984. Nagle...... 1980-1982. Hart ...... 1 979. Lind...... 1976-1979. Houck...... 1976. Schanes...... 1 974-1976. Skopic...... 1974.

1s Pressman, Jeffrey L. and Aaron B. Wildavsky. Implementation. Berkeley, Univ. of Cal. Press, 1973.

37-569 0 - 84 - 7 90 TABLE 2.-TENURE OF OFFICIALS RESPONSIBLE FOR ADMINISTERING ERISA-Continued

Position Official Tenure

Secretary of Labor...... Donovan...... 1981 topresent. Marshall...... 1977-1981. Usery...... 197 -1977. Dunlop...... 1975 1976. Brennan...... 1974-1975. Secretary of Treasury ...... Regan...... 1981 to present. Miller...... 1979-1981. Blumenthal ...... 1977-1979. Simon...... 1974-1977.

Effective followup is an ongoing part of the implementation proc- ess. Periodic assessments must be made so that a program can be evaluated. Evaluation can be both internal and external. Continual internal evaluation prevents problems from growing too large, but usually is done only when surplus resources exist. Firsthand knowledge of the agency permits a more informed appraisal, but obviously is less objective. External evaluations sometimes misin- terpret facts or miss salient factors of organizational behavior. Nev- ertheless, they are more objective and more powerful in bringing public pressure to bear on an agency. Because of the shortness of the transition period between the law's enactment and the effective dates, the honeymoon period was very short. Congressional committees, the General Accounting Office [GAO], various commissions, and the pension industry press quickly evaluated the performances of ERISA's implementors, both formally and informally, and found them to be lacking. 4 Public criticism undoubtably serves as an incentive for administrators to improve their operation because evaluation of ERISA prompted its implementors to improve their performances.

3. SPECIFIC IMPLEMENTATION PROBLEMS Scarcity of certain resources necessary for implementation trans- lates into specific programatic problems. The major implemenation problems shared by IRS and DOL during ERISA's initial stage in- cluded coping with overlapping jurisdiction, issuing regulations in a timely manner, processingexemptions, and paperwork. DOL also experienced difficulties filling staff positions. Although the PBGC surmounted its internal organizational problems handily, it faced policy problems when the time arrived to implement the Contin- gent Employer Liability Insurance sCEL ] and Multiemployer Pen- sion Plan Insurance provisions.

(A) OVERLAPPING DUAL ADMINISTRATION Congress' overlapping dual jurisdiction compromise severly threatened successful implementation of ERISA and exacerbated

14 A selected list of references appears at the end of this chapter. other problems. The process of writing regulations consumes a great deal of time under the best of circumstances. Proposed regu- lations must pass through a department's internal review, be pub- lished as proposed regulations in the Federal Register, be subject to amendment, and finally be published as regulations. The sheer number of regulations ERISA required and the short- ness of time before they were needed would have been a stumbling block in any event. However, with joint administration, it was nec- essary to send regulations back and forth between agencies, vastly increasing the number of decision points and time needed to get out proposed regulations. Coordinating policy with another agency always poses problems. Yet it is hard to imagine stranger bedfellows than DOL and IRS. Their differing missions and modus operandi contributed to diamet- rically opposed positions on some issues. The IRS, viewing its pri- mary role as that of a revenue raiser, considers the tax-qualified aspect of pensions first; a plan which does not comply with these guidelines can lose its qualified status. The preexisting routines and standard operating procedures within the department deter- mine how a new program will be implemented. The Labor Department concerns itself with the employee benefit issues of pensions. When considering pension regulations and policy, its primary consideration centers on the fulfillment of the pension promise, i.e., that workers will receive their benefits, sub- ject to ERISA's minimum standards. DOL's internal administrative procedures are considerably more flexible, but also more disorga- nized that those of the IRS. From these differing senses of mission and methods of operation, the IRS nd DOL had to coordinate their policies, all within a short period of time. One observer noted: The dual administration mandated by the statute is in- herently unworkable and has been recognized by most people who were involved in the original legislation as a necessary compromise, not a workable one.15 Widespread recognition of the intractability of overlapping dual jurisdiction resulted in President Carter's Reorganization Plan No. 4 which divided jurisdiction for major areas of ERISA between DOL and IRS. However, for 4 years, Labor and IRS struggled to work together to accomplish a complex task.

(B) PROHIBITED TRANSACTIONS By including prohibited transaction provisions in ERISA, Con- gress created a certain bottleneck for implementation. Since many pension abuses stemmed from breaches of fiduciary duty, reformers understandably wanted to prevent such action in the future. How- ever, they proposed to do so in an administratively maladroit

IS Testimony of Charles A. Moran, Senate Committee on Labor and Human Resources, "Over- sight of ERISA," pp. 412-13. manner. Instead of prohibiting transactions solely in cases in which plan members were harmed, ERISA prohibited a litany of transactions in all cases, but left a trap door open in the form of an exemption. Many of the transactions which ERISA prohibited for- merly constituted normal business practices. An often cited exam- ple of the over zealousness of the provision was the fact that it was a prohibited transaction for any employee of 30,000 corporations which had accounts with Prudential Insurance Co. to stay in the Washington Hilton because Prudential owned 50 percent of the hotel. To avoid violating these provisions, a fiduciary had to choose be- tween foregoing a potentially lucrative transaction or filing for an exemption. The exemption procedure required a case-by-case review of a financial transaction. Many pension plans had to wait years for decisions on their exemption applications, an extreme hardship since the opportunity for a business deal may pass quick- ly. DOL and IRS did not plan quickly enough for processing exemp- tion applications. The situation at Labor was especially bad; appli- cations for exemptions accumulated in piles in the office of the offi- cial charged with processing them without any action being taken. At the end of 1978, the average processing time for an exemption application was 15 months.1 6 Not surprisingly, public outrage grew and the GAO formally investigated the problem.

(C) PAPERWORK OVERKILL Paperwork was another major problem during ERISA's startup phase. Few legislators thought about the impact paperwork would have upon plan sponsors. They catered to the demands each profes- sional group made for having its own form and assumed that more paperwork equalled greater protection for plan participants." Plan sponsors, especially those with small plans, did not see it in the same way. At the very beginning, they experienced difficuties by not having the necessary forms early enough to design their data collection well; in fact, some filing deadlines approached with- out forms being ready. The enormity of the paperwork burden sunk in as deadlines for form submissions occurred. The requirement that annual reports be filed with both the IRS and DOL on different dates on different forms illustrates the lack of foresight which went into planning ERISA's paperwork. The Commission on Federal Paperwork selected ERISA as one of 18 Federal programs it would review. It studied problems associat- ed with multiple filing dates, diverse reporting requirements, em- ployees' need for disclosure, IRA's and Keoghs, and the means by which plans disseminate information. The Commission recommend- ed: eliminating unnecessary and duplicative reporting; providing plan participants with more understandable information; and sim-

16 U.S. Executive Office of the President. Office of Management and Budget. Study of the Ad- ministration of the Employee Retirement Income Security Act. Washington, U.S. Govt. Print. Off., January 1980. p. 37. '7 Interview with Michael Gordon. plifying the complex and confusing features of the program.1 " Since ERISA tried to clarify the information which was disclosed to plan participants, it is ironic that this was one of the criticisms of the Commission. The costs of complying with ERISA's paperwork provisions added insult to the injuries of having to fill out endless forms which were poorly designed. The Commission estimated cost savings of $74.5 million the first year and $283.2 million per year thereafter if its recommendations were adopted. 19 The burden of paperwork costs fell disproportionately on smaller plans, many of which had to hire outside experts to help them fill out forms. The Commission on Federal Paperwork was but one voice in the rising chorus against ERISA's paperwork "overkill." Slowly DOL, IRS, and the PBGC began to consolidate their forms and to elimi- nate some reporting requirements. Although plan sponsors still want to simplify paperwork, their burden has been reduced signifi- cantly. Despite improvements, many problems associated with ERISA paperwork continue. The costs of filing and processing paperwork remain high, both for plan sponsors and the agencies. Agencies which have ceded their authority over paperwork often find that the information forwarded to them by the collecting agency is in- complete, inappropriate for their purposes, or late. Extensive dis- closure of plan information has not had the desired effect of reas- suring plan participants; many worry about the security of their plans when they receive incomprehensible information about it.

(D) DOL STAFFING PROBLEMS Assembling a qualified staff provided to be problematic for DOL. It often fails to fill all of its allocated staff positions. The GAO cited Labor's delay in hiring sufficient qualified staff as a "primary cause of slow progress in issuing regulations and processing exemp- tion applications". 2 0 Staffing problems continue today in Labor's enforcement program. The Department simultaneously claims that it needs more staff positions to put teeth in its programs, but admits failing to fill all of the positions it has. 2 '

(E) PBGC: CELI AND MULTIEMPLOYER PLANS After an initial heady period of organization, the PBGC came face-to-face with program responsibilities which proved to be more intractable than mailing out 100,000 premium notices less than a month after operations began. Upon examination, the PBGC found that implementing the contingent employer liability insurance and the multiemployer termination insurance provisions might jeopard- ize the financial soundness of the corporation. Since the conceptual details of both of these programs had not been worked out, the

15 U.S. Commission on Federal Paperwork. The Employee Retirement Income Security Act. Washington, Dec. 3, 1976, p. 1 19 Ibid. 20 U.S. General Accounting Office. Efforts to Implement the Employee Retirement Income Se- curity Act of 1974 by the Department of Labor. HRD-77-99, July 6, 1977. Washington, 1977, p. 16. Unpublished briefing materials prepared by PWBP for the National Pension Forum of the ERISA Advisory Council. PBGC had to develop policy before administration could even begin. ERISA charged the PBGC with the responsibility of finding pri- vate insurers to develop a contingent employer liability insurance program or of doing so itself in response to intense opposition to the concept of employer liability up to 30 percent of net worth in the case of plan termination. When the PBGC's effort to induce pri- vate insurers to provide CELI failed, it tried to develop a program on its own. However, it soon realized the unsoundness of trying to provide insurance against events that can be controlled by the in- sured. In testimony before the Senate Subcommittee on Labor, PBGC Executive Director Matthew Lind stated: A CELI program which provides broad relief from em- ployer liability appears to be financially unsound and, more importantly, could undermine the whole termination insurance scheme by driving premiums to intolerable levels through inducing and inviting terminations. 2 2 Congress apparently agreed with Lind's assessment since CELI was revoked. Implementing the Multiemployer Termination Insurance Pro- gram proved to be more problematic than anyone had anticipated. Congress did not know very much about the multiemployer plan universe when it enacted ERISA. It accepted the assurances of union plan sponsors that they are relatively healthy. However, two important considerations were overlooked: the relative youth of the plans and the strong economy during the period examined. Both conditions make plans appear healthier than they are because they have a high ratio of participants to beneficiaries and many em- ployed participants contribute to the plan. Persons responsible for setting the January 1, 1978 effective date for Multiemployer Termination Insurance probably had an inkling that problems might arise. However, no one anticipated the breadth and scope of problems which existed in the multiemployer plan universe or how ERISA would make those problems worse. Problems ranged from declining industries to unfunded or under- funded plans with sponsors withdrawing so that they would not be left holding the bag. ERISA prohibited decreasing benefits, the usual remedy for alleviating financial problems. The PBGC, faced with financial insolvency if multiemployer plans started to terminate, succeeded in having the deadline for terminations of multiemployer plans extended twice, the actual ef- fective date being May 1, 1980. In the meantime, it lobbied Con- gress for the passage of the Multiemployer Pension Plan Amend- ments Act of 1980 [MPPAA] which effectively changed the rules of the game for multiemployer plan terminations. MPPAA made plan insolvency the only insurable event and assessed employer liability for sponsors withdrawing from the plan. A retroactive effective date of April 29, 1980 for withdrawing employers was set to pre- vent massive withdrawals before the law became effective. Al- though heartily disliked by all, MPPAA seems to be surviving nu-

22 Senate Committee on Labor and Human Resources. Oversight of ERISA. pp. 498-9. merous court challenges and has probably saved multiemployer plans and the PBGC. E. STAGE 2: "SETTLED DUST" OF NORMAL IMPLEMENTATION It usually takes several years to pass from the startup phase to the initial phase of full implementation. Features characteristic of this second stage include: the presence of routines and standard op- erating procedures; program enforcement becomes more important; litigation begins; all major regulations have been issued; and amendments to the law are considered. If chaos and anxiety domi- nate the startup phase of implementation, relief prevails during the initial part of this second stage. In general, most people, pleased that the program finally is working, do not want to consid- er making any major changes which would require another adjust- ment phase. When the relief at having the program operational fades, satis- faction with the overall operation of the program often declines. Major changes might be proposed; public support for them might grow. If one were to single out critical events in the implementation of ERISA, the passage of MPPAA and President Carter's Reorganiza- tion Plan No. 4 could be cited as major turning points since these events ended the startup phases of implementation. 1. REORGANIZATION PLAN No. 4 As discussed above, overlapping dual jurisdiction impeded the implementation efforts of both IRS and DOL. Each lacked auton- omy to carry out its mandate, shackled together as they were. The inherent unworkabilify of shared jurisdiction was evident to all by 1977. IRS and DOL began to exchange a series of memos of under- standing in that year to separate the threads of the program into two distinct piles. Their negotiations closely mirrored the original division suggested in the Senate bill. The final metamorphosis of these discussions came in the form of Reorganization Plan No. 4. Effective on December 31, 1978, President Carter's Reorganiza- tion Plan No. 4 marks the beginning of ERISA's normal phase of implementation. The primary features of the plan include: (1) Transferring authority over minimum standards in par- ticipation, vesting, and funding from Labor to Treasury. (2) Transferring authority for fiduciary obligations, including prohibited transactions, from Treasury to Labor. (3) Continuing coordination on notices of intent to disqualify multiemployer plans. (4) Continuing each agency's authority over enforcement. (5) Evaluation of the program, with long-term recommenda- tions by January 31, 1980. Thus, Reorganization Plan No. 4 delegated responsibility for fidu- ciary and disclosure matters to Labor and benefits issues (mini- mum standards) to Treasury. Although many persons concerned with pension rights for plan participants and beneficiaries dislike this division of tasks, citing the IRS's overriding concern for raising revenue, it was a compromise necessary for reorganization. 2 3 While the plan divided most of the responsibilities for administer- ing ERISA between IRS and DOL, both retained authority over en- forcement activities, necessitating the need for continued coordina- tion in this area. Since the reorganization did not add any new pro- gram responsibilities and eliminated others, DOL and IRS imple- mented it without major problems.

2. ENFORCEMENT By 1977 and 1978, Labor and Treasury became involved in en- forcing ERISA and in litigating it. Program enforcement follows other administrative tasks since rules, regulations, guidelines, and operating procedures defining appropriate behavior must be in place before violations can be assessed. Program enforcement consists of activities an agency undertakes to insure voluntary or involuntary compliance with a program. Im- plementors prefer voluntary compliance because it is much cheap- er, quicker, and less confrontational than involuntary compliance. The essence of voluntary compliance is informing the public of the need to comply and the way to do it. Methods of obtaining volun- tary compliance range from running informational hotlines to pub- licizing a program to permitting a plan to come into compliance without filing a suit. To enforce compliance involuntarily, an agency must develop an enforcement strategy. The first step is to determine who is not complying. Depending upon the resources available, audits can be done of all or some of the parties. If only a sample is to be audited, key choices about the selection of the parties to be audited need to be made. Choices range from strictly random audits to audits of specific categories of plans-eg. large plans or union plans.

(A) IRS Initially, IRS concentrated its enforcement efforts on examining compliance with participation and vesting standards, almost to the exclusion of funding and disclosure standards. Most of the staff's time was spent issuing determination letters rather than enforcing involuntary compliance. At the very beginning, it also set up ERISA hotlines to answer questions. Subsequent enforcement of ERISA did not prove to be an entire- ly smooth road for the IRS. In 1979, the GAO criticized the Service for selectively enforcing certain parts of the law to the detriment of others, and for not being more aggressive in seeking out plans in noncompliance. 2 4 At the time, IRS and DOL had plans to create an automated system which would review annual reports and select out those which had the strongest potential for noncompliance. Al- though the IRS examined a sample of 25,000 plans to form a factu- al base for developing criteria for its targeting system, no system has emerged to date.

23 Interviews with Karen Ferguson and Ian Lanoff. 24 U.S. General Accounting Office. Internal Revenue Service Efforts and Plans to Enforce the Employee Retirement Income Security Act. HRD-79-55, March 28, 1979. Washington, 1979. In late 1980, IRS instituted the ERISA Non-Compliance Enforce- ment Program [ENCEP]. The program seeks to find plans which did not file for new determination letters after ERISA's enactment. The plan aims to keep plans qualified so that participants do not lose benefits.

(B) DOL Labor's enforcement efforts were somewhat different. In 1975, it undertook an investigation of the Central State Teamsters plan- an activity which would prove to be both protracted and costly in terms of available resources. Exept for this investigation, before 1978, most of DOL's field personnel provided general and technical information about ERISA rather than devoting time to enforce- ment. Lack of direction from Washington left field personnel at loose ends. The GAO issued reports critical of enforcement efforts in both agencies. Shortly after the Department of Labor was brought to task in July 1977, it announced priorities in its enforcement pro- gram. 25 Under the leadership of PWBP Administrator Ian Lanoff, DOL decided to implement an enforcement policy based on the sig- nificant case theory-each region and area would select plans to audit from among the largest plans in the region. By decentralized programs in the field, disparities arose about significant cases be- tween regions. For example, much smaller plans constituted signifi- cant cases outside of major industrial areas. The selection from among the largest plans troubled some. While trying to obtain the greatest good for the greatest number by focus- ing on large plans can be defended on the grounds of pragmatism, others claimed that the most serious defects were to be found in smaller plans and that participants of those plans should be ac- corded the same protection as participants of larger plans. Once the significant case enforcement plan got off the ground, field inspectors succeeded in generating a lot of good cases for DOL's Office of the Solicitor. In fact, field inspectors grew frustrat- ed as their cases became backlogged in the Solicitor's office, a likely outcome when a program must compete with other programs within a department. Unfortunately, enforcement programs in DOL seem to survive only as long as the PWBP administrator who inaugurated it. Once the Reagan administration appointee took office, the significant case approach was abandoned. The new administrator did not im- plement an alternative enforcement program, letting the program drift for 2 years; often, DOL's enforcement office suffered because of vacancies in key positions. Within the Department, the emphasis shifted from civil to criminal enforcement. The Solicitor's Office, formerly overburdened with ERISA cases, did not have many cases on the books. Since Robert Monks took over as ERISA administrator in Janu- ary 1984, a new approach to enforcement has been announced. Cen- tralized enforcement similar to that used by the Securities and Ex-

25 See, e.g. U.S. General Accounting Office. Efforts to Implement ERISA. 98 change Commission will replace the decentralized field approach. However, that plan has yet to be fully implemented.

3. LINGERING PROBLEMS Although ERISA enforcement constitutes the weakest link in im- plementation, other problems remain. DOL experiences delays in receiving information from the IRS. It also lacks funding for re- search and statistics, forcing it to rely on outdated information and data which does not suit its purposes. The Department of Health and Human Services currently funds more pension research than DOL does. DOL still has not promulgated all of its initial regulations. As of May 31, 1984, 39 regulations remain to be issued. Backups in the exemption process also are reoccurring. In the past 5 years, applications have risen 100 percent. However, many qualified staff members have left the agency without being re- placed, and class exemptions have been issued so a large number of exemptions cannot be closed at once. The average processing time currently stands at 8 months and is expected to rise in the future. 26

4. EVALUATING THE IMPLEMENTATION OF ERISA Despite these problems, 10 years after its enactment, ERISA im- plementation has reached a plateau of satisfactory administration; the initial implementation problems have been eliminated or alle- viated. The IRS and PBGC retain their initial administrative struc- tures with minor changes. DOL continues to reorganize frequently, its most recent change occurring in May 1984 when PWBP was again removed from LMSA. However, its internal changes no longer prevent it from carrying out its responsibilities in a satisfac- tory manner. DOL, IRS, and the PBGC administer ERISA in a competent, if uninspired manner. The program has been most stable at the IRS, in part because its leadership has been remarkably continuous for the past 10 years. Although the program has not changed much for DOL, internal reorganizations within the department and a succes- sion of seven leaders in 10 years made implementation less smooth. The PBGC has had several major legislative battles, many chal- lenges to its authority in court, and a succession of leaders. Howev- er, it has been able to survive these difficulties and copes with its smaller mandate. ERISA had to be adapted in many ways by the agencies, the President, and Congress to make it work. To obtain a workable ad- ministrative apparatus, the overlapping dual administration fea- ture of ERISA needed to be struck down. President Carter's Reor- ganization Plan No. 4 delivered the coup de grace by dividing the overlapping duties between Labor and Treasury, except in the area of enforcement which continues to be coordinated loosely. The agencies reduced the unwieldy paperwork burden substan- tially by reducing the amount of required reporting and becoming more efficient in handling the work load. They have accomplished

28 Unpublished briefing materials of the National Pension Forum. this by sharing information, eliminating unnecessary forms, simpli- fying other forms, and lengthening the interval of time for filing certain forms. While plan sponsors would like to see the reporting and disclosure requirements pared down even more, many persons express concern about preserving ERISA's purpose-the protection of pension benefits. A balance must be maintained between mini- mizing paperwork and safeguarding pension security. Time and experience cured many of the initial implementation problems of Labor and IRS. Experience with processing reports and exemptions expedited those processes. The backlog of unpublished regulations has shrunk and should continue to decline. 2 7 These achievements reflect internal adaptations and cooperative agree- ments reached by the agencies. The PBGC faced problems which jeopardized its financial solven- cy and which needed to be resolved by legislation rather than ad- ministrative or executive action. CELI had to be written out of the law because of its inherent unworkableness. The beginning dates for terminations under the Multiemployer Termination Insurance program had to be extended by legislation to stave off massive ter- minations. Premium rates had to be raised for both single and mul- tiemployers. MPPAA had to be enacted to change the multiemploy- er termination insurance program. At present, the Corporation is still awaiting the passage of single employer legislation designed to remove incentives which induce employers to terminate underfund- ed pension plans, dumping unfunded liabilities on PBGC, and to raise the premium rate for single employer termination insurance. While the agencies which administer ERISA are doing a credible job holding down the fort, one is left with the general impression that such is the extent of their current activity. F. STAGE 3: MATURE IMPLEMENTATION Part of the problem with current ERISA administration is that the agencies which administer it have an oligopolistic relationship. None is supreme; each has its sphere of influence. None could take a leadership position without eclipsing the others. Thus, ERISA im- plementation will never Unove into a mature stage of implementa- tion under the current administrative division. As noted previously, an agency reaches the mature phase of im- plementation when it has assumed a leadership position in its policy area. The public recognizes it as the leading authority in its field whose expertise usually is backed up by a strong research and statistics component within the organization. Mature agencies ex- ercise their leadership both actively and passively. They might lead the charge by pushing Congress to make certain policy changes in their area. Or, a more passive agency might respond only when asked for its opinion; however, it would react in a clear and timely manner. An agency in the mature phase of implementation has two im- portant sources of power: expertise and a strong constituency. The agency uses these resources to transcend mere program adminis- tration; it affects the policy outcome.

27 Ibid. Appendix I. Some agencies never reach the mature phase of implementation. Their function might not be suitable for policymaking (e.g., a motor vehicle administration) or their administrators might never possess the will and the resources necessary to play a leading role in pol- icymaking. The type of administrative structure an agency has and its position within the Federal bureaucracy also determine whether or not it can assume a position of power. 2 8 Whether it is desirable for an agency to achieve a mature state of implementation is a policy question. Given changes in laws which affect who gets a pension under what circumstances, the function clearly is suitable for policymaking at the agency level in ERISA's case. A strong case can be made for the merits of having an agency take an active part in pension policy because of the im- portance of pension funds to retirees and capital markets. Howev- er, it is difficult to foresee ERISA implementation attaining a mature stage without consolidation of the tripartite jurisdiction. Consolidating ERISA administration into a single agency would cut down on some of the areas of the program which still run less than smoothly. As table 3 shows, the costs of administering ERISA over the past 10 years have exceeded $800 million. Undoubtably, some savings could be achieved by eliminating duplicative adminis- trative structures and the need to coordinate. Taking ERISA out of DOL probably would save a lot of money in and of itself. Former PWBP Administrator Ian Lanoff noted wryly: So much time and money is spent arguing whether it (ERISA administration) should be under LMSA. It's ridicu- lous. It's a waste of everyone's time; you probably could staff up and run an agency for all it's cost.2 9

TABLE 3.-FUNDS EXPENDED FOR ADMINISTERING ERISA, FISCAL YEARS 1975-85 [inthousands] Fiscalyear Labor(PWBP) PBGC IRS(EP)

1975 ...... 3,194 2,888 22,752 1976 ...... 1,942 7,253 31,203 TO' ...... 4,585 2,83 1 8,377 1977 ...... 18,585 13,224 35,880 1978 ...... 22,844 14,550 39,402 1979 ...... 27,63 1 16,429 39,769 1980 ...... 28,699 19,908 40,514 1981 ...... 32, 83 21,499 41,405 1982 ...... 25,441 22,911 43,400 1983 ...... 26,119 27,88 1 45,000 1984 (estim ated)...... 29,200 31,516 56,400 1985 (estim ated)...... 29,299 32,898 62,600 Total(estim ated)...... 259,622 213,788 427,300 Transitionquarter due to change in fiscal year. Sources:Bureau ofthe Budget, Office ofManagement andBudget, Internal Revenue Service, theBudget ofthe U.S. Government However, whether or not ERISA administration will be consoli- dated into a single agency is a question open for the future. The 1980 OMB task force which reviewed Reorganization Plan No. 4

2" Seidman, Harold. Politics, Position, and Power. New York, Oxford Univ. Press, 1980. 21 Interview with Ian Lanoff. noted that the reorganization was working to alleviate many of ERISA's initial problems, but that the solution was merely a tem- porary one. However, they did recommend that no consolidation be made at that time.30 In the past 2 years, sentiment in the pension community con- cerning a single agency apparently shifted.3 1 Persons who ex- pressed satisfaction with the status quo 2 years ago advocate con- solidation today. Persons who expressed opposition 2 years ago are more likely to withhold judgment today, noting that no bills on consolidation have emerged from Congress. Many obstacles stand in the way of achieving a single agency. The structure and leadership of a new agency would be open to much debate and possible disagreement which could kill the idea entirely. It might not be possible to wrestle pensions away from the IRS entirely, since the issue remains very much a tax issue. Also, it is critical that interest be aroused on the Hill before anything can be accomplished. Few if any legislators have appeared to accept the mantle of leadership from ERISA's founding fathers. Critics of ERISA implementation acknowledge that current prob- lems of administration are relatively minor, especially when com- pared with the problems of the past. Yes, enforcement, exemptions, and paperwork could be handled better. However, the major sins are those of omission, rather then commission. Members of the pension community feel that their issue is one of vital importance to the nation. Despite its importance to the mil- lions of Americans who rely and will rely on pensions for retire- ment income, and the importance of those funds for capital mar- kets, the issue lacks a sufficient number of powerful advocates in the Congress or in the executive branch. Funds for research remain inadequate. Program implementors do not look into the future. Members of the pension community desire a strong advo- cate for pensions and many believe that a single agency could be that advocate. G. CONCLUSION In the past 10 years, implementation of ERISA in the executive branch has passed from the chaos of starting up a program to the stable routine of normal implementation. Although many problems originated in the legislation itself, the agencies charged with ad- ministering ERISA adapted themselves or were adapted by execu- tive action or congressional amendment. Because ERISA's administration remains divided between three agencies, no preeminent advocate for private pensions exists today. ERISA implementation has not reached the mature stage in which the administrative agency assumes an active policy role. Whether ERISA implementation will ever reach the mature phase of implementation remains unclear. The status quo repre- sents competent, if uninspired administration. Whether this stand- ard will remain acceptable is a policy question which must be an-

30 U.S. Office of Management and Budget. Study of the Administration of ERISA. pp. 75-8. 31 Author's experience based on difference between interviews conducted in 1982 and 1984. swered before attempts to alter ERISA's administrative apparatus will be made. H. BIBLIOGRAPHY Dahl, Robert A. "The Science of Public Administration: Three Problems." Public Administration Review 7 (1977) pp. 1-11. Edwards, George C., III and Ira Sharkansky. "The Policy Predica- ment and Implementing Public Policy." San Francisco: W.H. Freeman and Co., 1978. Goodnow, Frank. "Politics and Administration." New York: Mac- millan, 1900. National Pension Forum of the ERISA Advisory Council. Briefing material for inaugural meeting, June 14, 1984. Pressman, Jeffrey L. and Aaron B. Wildavsky. "Implementation." Berkeley: University of California Press, 1973. Seidman, Harold, "Politics, Position, and Power." 3d edition. New York: Oxford University Press, 1980. Simon, Herbert A. "Administrative Behavior: A Study of Decision- making Processes in the Administrative Organization," 2d edi- tion. New York: Macmillan, 1961. U.S. Commission on Federal Paperwork. "The Employee Retire- ment Income Security Act." Washington, D.C.: U.S. Govern- ment Printing Office, 1976. U.S. Congress, House, Subcommittee on Labor Standards of the Committee on Education and Labor. "Oversight on the Em- ployee Retirement Income Security Act of 1974." Washington, D.C.: U.S. Government Printing Office, 1975. U.S. Congress, Senate, Subcommittee on Labor of the Committee on Human Resources, "Oversight on ERISA," 1977. Washing- ton, D.C.: U.S. Government Printing Office, 1977. U.S. Executive Office of the President, Office of Management and Budget. "Study of the Administration of the Employee Retire- ment Income Security Act." Washington, D.C.: Office of Man- agement and Budget, January 1980. U.S. General Accounting Office, "Efforts to Implement the Employ- ee Retirement Income Security Act of 1974 by the Department of Labor." Washington, D.C.: U.S. General Accounting Office, July 6, 1977. U.S. General Accounting Office, "Internal Revenue Service Efforts and Plans to Enforce the Employee Retirement Income Securi- ty Act." Washington, D.C.: U.S. General Accounting Office, March 28, 1979. Wilson, Woodrow. "The Study of Public Administration." Political Science Quarterly, December 1941, pp. 486-506. I. SELECTED REFERENCES ON ERISA LEGISLATIVE OVERSIGHT: 1974-84 This supplemental bibliography lists materials concerning over- sight, primarily in the legislative branch, of the implementation and administration of the Employee Retirement Income Security Act of 1974 [ERISA] from 1974-84. Oversight hearings on specific amendments or conducted during the appropriation process, are not included. The citations are alphabetically arranged according to the name of the congressional committee or oversight agency, and then by date. References were compiled by the Congressional Research Service from the bibliographic database maintained by the Library Serv- ices Division of the Congressional Research Service and the Library of Congress Computerized Catalog. Additional citations were pro- vided from the CISID Federal evaluations file of the General Ac- counting Office. U.S. Congress. Senate Conference Committee, 1974. Employee Re- tirement Income Security Act of 1974; conference report to ac- company H.R. 2, 93rd Congress, 2nd session. Washington, GPO, 1974. 387 p. U.S. Congress. House. Committee on Education and Labor. Pension Benefit Guaranty Corporation Insurance Premiums; Report to accompany H. Con. Res. 369, 95th Congress, 1st session. Wash- ington, GPO, 1977. 9 p. --- , Oversight Hearings on ERISA Reporting and Disclosure Re- quirements. Hearings, 97th Congress, 1st and 2nd sessions, Nov. 16, Dec. 9, 1981; Jan. 29, 1982. Washington, GPO, 1983. 460 p. Committee on Education and Labor. Subcommittee on Labor Standards. Oversight of the Employee Retirement Income Security Act of 1974. Hearings, 94th Congress, 1st ses- sion, Apr. 29-May 6, 1975. Washington, GPO, 1975. 981 p. Joint Oversight Hearing on the Employee Retirement Income Security Act of 1974. Joint hearings before the Subcom- mittee on Labor Standards of the Committee on Education and Labor and the Subcommittee on Oversight of the Committee on Ways and Means, House of Representatives, 94th Congress, 1st session, Dec. 9, 1975. Washington, GPO, 1976. 125 p. Oversight on Department of Labor Administration of ERISA. Hearing, 94th Congress, 1st session, Feb. 4, 1976. Wash- ington, GPO, 1976. 33 p. --- , General Oversight on ERISA. Hearing, 94th Congress, 2nd session, June 1 and July 20, 1976. Washington, GPO, 1976. --- , Employee Benefit Administration Act of 1977. Hearings, 95th Congress, 1st session, on H.R. 4340, Mar. 17 and Apr. 20, 1977. Washington, GPO, 1977. 237 p. --- , Oversight on ERISA, 1978. Hearings, 94th Congress, 2nd session, on Public Law 93-406, May 17, 1978 and June 1, 1978. Washington, GPO, 1978. 1306 p. Committee on Education and Labor. Subcommittee on Labor Standards. Pension Task Force. ERISA oversight report. Washington, GPO, 1977. 11 p. Committee on Government Operations. Reorganization Plan Number 4 (ERISA Administration); Report to accompany H. Res. 1308, including cost estimate of the Congressional Budget Office, 95th Congress, 2nd session. Washington, GPO, 1978. 31 p. --- , Committee on Government Operations. Legislation and Na- tional Security Subcommittee. Reorganization Plan No. 4 of 1978 (Employee Retirement Income Security Act). Hearing, 95th Congress, 2nd session, Sept. 12, 1978. Washington, GPO, 1978. 74 p. --- , Committee on Ways and Means. Subcommittee on Over- sight. Pension Reform Act of 1974. Hearings, 94th Congress, 1st session, Nov. 20, 1975. Washington, GPO, 1975. 79 p. --- , Employee Retirement Income Security Act of 1974. Field hearing, 94th Congress, 2nd session, May 24, 1976. Washington, GPO, 1976. 109 p. --- , Teamsters' Central States Pension Fund and General ERISA enforcement. Hearings, 95th Congress, 1st session, Mar. 14-July 20, 1977. Washington, GPO, 1977. 2 v. (931 p.) --- , Central States Teamsters Fund. Hearing, 95th Congress, 2nd session, Mar. 22, 1978. Washington, GPO, 1978. 209 p. --- , Pension Plan Termination Insurance Issues. Hearing, 95th Congress, 2nd session, Sept. 28, 1978. Washington, GPO, 1978. 87 p. --- , Teamster's Central States Pension Fund Reform and ERISA Enforcement Remedies. Hearing, 97th Congress, 2nd session, July 26, 1982. Washington, GPO, 1982. 252 p. --- , Select Committee on Aging. Pension Asset Raids. Hearing, 98th Congress, 1st session, Sept. 28, 1983. Washington, GPO, 1984. 293 p. --- , Select Committee on Aging. Subcommittee on Retirement Income and Employment. National Pension Policies: Private Pension Plans. Hearings, 95th Congress, 2nd session Feb. 27, Mar. 13, and Apr. 6, 1978. Washington, GPO, 1978. 735 p. U.S. Congress. Senate. Committee on Finance. Pension Require- ments Simplification Act; report to accompany S. 2352, 93rd Congress, 1st session. Washington, GPO, 1977. 23 p. --- , Committee on Finance. Subcommittee on Private Pension Plans. Paperwork requirements of the Pension Reform Act of 1974. Joint hearings before the Subcommittee on Private Pen- sion Plans and the Subcommittee on Financial Markets of the Committee on Finance and the Select Committee on Small Business, United States Senate, 94th Congress, 2nd session, Feb. 2 and 3, 1976. Washington, GPO, 1976. 166 p. --- , Committee on Finance. Subcommittee on Private Pension Plans and Employee Fringe Benefits. Pension Simplification and Investment rules. Joint hearings before the Subcommittee on Private Pension Plans and Employee Fringe Benefits of the Committee on Finance and the Select Committee on Small Business, United States Senate, 95th Congress, 1st session May 10-July 18, 1977. Washington, GPO, 1977. 570 p. --- , Pension Plan Simplification. Hearing, 95th Congress, 2nd session, June 27, 1978. Washington, GPO, 1978. 105 p. --- , Committee on Governmental Affairs. Reorganization Plan No. 4 of 1978. Hearing, 95th Congress, 2nd session, on a reorga- nization plan to improve the administration of the Employ- ment Retirement Income Security Act of 1974, prusuant to 5 U.S.C. 903 (91 Stat. 30), Sept. 7, 1978. Washington, GPO, 1978. 54 p. -, Reorganization Plan No. 4 of 1978; report to accompany S. Res. 537. Washington, GPO, 1978. 150 p. --- , Committee on Human Resources. Subcommittee on Labor. Oversight of ERISA, 1977. Hearings, 95th Congress, 1st session, Oct. 11, 13, and 18, 1977. Washington, GPO, 1978. 816 p. -- , ERISA Improvements Act of 1978. Joint hearings before the Subcommittee on Labor of the Committee on Human Re- sources and the Subcommittee on Private Pension Plans and Employee Fringe Benefits of the Committee on Finance, United States Senate, 95th Congress, 2nd session, Aug. 15-17, 1978. Washington, GPO, 1978. 1461 p. --- , Committee on Labor and Human Resources. ERISA Im- provement Act of 1979. Hearings, 96th Congress, 1st session, Feb. 6-8, 1979. Washington, GPO, 1979. 1416 p. --- , ERISA Simplification Act of 1979. Hearing, 96th Congress, 1st session, Nov. 7, 1979. Washington, GPO, 1980. 169 p. --- , Committee on Labor and Human Resources. Subcommittee on Labor. Retirement Income Incentives and Administrative Simplification Act of 1981. Hearings, 97th Congress, 1st session, Nov. 4, 1981-Jan. 26, 1982. Washington, GPO, 1982. 1057 p. --- , Oversight of the Pension Benefit Guaranty Corporation, 1982. Hearings, 97th Congress, 2nd session, May 19, 1982. Washington, GPO, 1982. 192 p. -- , Oversight of Private Pension Plans, 1983. Hearings, 98th Congress, 1st session, Washington, GPO, 1983. 351 p. --- , Committee on Labor and Public Welfare. Subcommittee on Labor. Legislative History of the Employee Retirement Income Security Act of 1974, 94th Congress, 2nd session. Washington, GPO, 1976, 3 v. --- , Committee on Small Business. Subcommittee on Productivi- ty and Competition. The Effects Which ERISA and MPPAA [Multiemployer Pension Plan Amendments Act] Have Had on the Competitiveness of Small Business. Hearing, 97th Con- gress, 1st session. Aug. 14, 1981. Washington, GPO, 1981. 384 p. U.S. Department of Labor. Administration of the Employee Retire- ment Income Security Act Through December 31, 1975; report of the Secretary of Labor to the Congress. Washington, GPO, 1976. 45 p. U.S. General Accounting Office. Efforts to Implement the Employ- ee Retirement Income Security Act of 1974 by the Department of Labor; report to the Senate Committee on Human Resources by the Comptroller General of the United States. Washington, 1977, HRD-77-99, July 6, 1977, 20 p. --- , Examination of Financial Statements, Pension Benefit Guaranty Corporation, Fiscal Year 1977; communication from the Comptroller General of the United States, 96th Congress, 1st session. Washington, GPO, 1979. 16 p. --- , Tax Revenues Lost and Beneficiaries Inadequately Protect- ed When Private Pension Plans Liquidate; report by the Comp- troller General of the United States. Washington, 1981, HRD- 81-117, 52 p. -- , Better Management of Private Pension Plan Data Can Reduce Costs and Improve ERISA Administration; report by the Comptroller General of the United States. Washington, 1981, HRD-82-12, 118 p. -, IRS Needs to Curb Excessive Deductions for Self-employ- ment Retirement Plans; report to the Commissioner, Internal Revenue Service. Washington, 1982. GGD-82-85, 17 p.

37-569 0 - 84 - 8 106

--- , Legislative Changes Needed to Financially Strengthen Single Employer Plan Insurance Program; report by the Comp- troller General of the United States. Washington, 1983, HRD- 84-5, 73 p. --- , Review of the Pension Benefit Guaranty Corporation's Progress Towards Improving Accounting and Internal Weak- nesses. Washington, 1983, AFMD-84-2, 4 p. --- , Coordination Between the Departments of Labor and Jus- tice in Investigating Criminal Activities of Labor Unions and Employee Pension and Welfare Benefit Funds. Washington, 1984, HRD-84-9, 17 p. --- , Improved Processes Needed to Insure Timely Benefit Pay- ments to Participants of Single Employer Pension Plans; report to the Executive Director, Pension Benefit Guaranty Corporation. Washington, 1984. HRD-84-58, 29 p. U.S. Office of Management and Budget. Administration of the Em- ployee Retirement Income Security Act, ERISA, a report to Congress. Washington, 1980. 81 p. U.S. President, 1977-1981 (Carter). Amendment to Reorganization Plan No. 4 of 1978; message, 95th Congress, 2nd session. Wash- ington, GPO, 1978. 3 p. Chapter 5 WHAT IMPACT HAS ERISA HAD ON DIFFERENT TYPES OF PENSION PLANS? (Prepared by Dallas L. Salisbury,' President, Employee Benefit Research Institute, Washington, DC) A. INTRODUCTION Economic security in retirement has been a national policy goal since enactment of the Revenue Act of 1921 and the Social Security program in 1935. Retirees presently seek to meet this objective through a combination of government-provided, employer-provided, and personal sources: Social Security, medicare, employer pensions, and other employer-paid capital accumulation programs, employer- provided post-retirement health insurance, personal savings, family transfers, and employment. Employers sponsor two basic types of retirement plans: 2 Defined benefit plans provide a clearly stated retirement benefit with a fluctuating annual contribution. The risk/ reward of investment performance rests with the employer sponsor. A small number of these plans will pay a lump sum upon retirement but most provide a monthly benefit check. Defined contributionplans provide a specified annual contri- bution with a final benefit based on contributions and investi- ment returns. The risk/reward of investment performance rests with the individual worker. Most of these plans pay out the account balance as a lump sum at the point of retirement or job termination. Some DC plans, such as TIAA/CREF, do pay benefits as a monthly annuity. The Employee Retirement Income Security Act of 1974 [ERISA] was enacted to assure that retirement plans be soundly funded and fully communicated (previous chapters discuss ERISA purposes in detail). When ERISA was enacted, defined benefit pension plans were a majority of the pension universe. It was the employers' principle means of contributing to economic security in retirement beyond Social Security. As noted in chapter 1, the desire to

' The views expressed in this chapter are those of the author and should not be attributed to the Employee Benefit Research Institute, its staff, officers, directors, or sponsors. EBRI is a non- profit, nonpartisan, public policy research organization. Before joining EBRI, Mr. Salisbury served in senior career policy research positions at the U.S. Department of Labor and the U.S. Pension Benefit Guaranty Corporation. 2 A full discussion of the various types of plans and the relative advantages and disadvantages of each can be found in two books published by the Employee Benefit Research Institute. Funda- mentals of Employee Benefit Programs. Washington, EBRI, 1983. This book offers basic descrip- tions. Economic survival in Retirement: Which Pension Is For You? Washington, EBRI, 1982. This book offers analysis and discussion. "secure" defined benefit plans motivated most of ERISA's provi- sions. When ERISA was passed, there were approximately 425,000 "pension" plans in the United States. Today there are over 820,000 public and private employer-sponsored plans covering more than 55 million total workers and 50 million nonagricultural wage and salary workers. Private employer plans cover 36.5 million nonagri- cultural workers, 51 percent, and 27.6 million workers who meet ERISA minimum participation standards, or 60.1 percent. Also, the post-ERISA environment includes over 16 million individual retire- ment accounts [IRA's]. While ERISA made many changes that affected the cost of main- taining plans, it did not change the three fundamental aspects of the law that have most readily encouraged plan provision: -First, that negotiating the. provision of pensions can be a part of the collective bargaining process. -Second, that contributions to plans are tax deductible expenses for employers; and -Third, that workers' taxes on both contributions and invest- ment earnings are deferred until retirement or until a distri- bution is received from the plan. Additionally, it extended the advantages of tax deferral to indi- viduals by establishing the IRA. Since enactment of ERISA in 1974, the universe of retirement savings arrangements has become more diverse. And, understand- ing what a "pension plan" is has become more difficult. B. WHAT IS A PENSION PLAN UNDER ERISA? ERISA muddied the definitional waters of employee benefits. Pre-ERISA the Webster's definition of a pension plan held sway: * * * 1. a payment, not wages, made regularly to a person (or to his family) who has fulfilled certain condi- tions of service, reached a certain age, etc. Only a program that paid benefits in the form of an annuity was a pension-such as Social Security. In the employer world this meant nearly all defined benefit plans, the TIAA-CREF defined contribution with annuity payout plan, and some other defined contribution plans. In other words, a "pension" plan was one that assured a stream of payments to an individual during their retirement years to maintain a level of economic security. The pre-ERISA world had three other primary types of "capital accumulation" plans that are now covered by ERISA: -Money purchase defined contribution plans into which the em- ployer paid a fixed percentage of pay into an individual ac- count for employees which would be paid to vested employees upon termination of employment (retirement or otherwise) in a lump sum. -Profit-sharing defined contribution plans into which the em- ployer paid some percentage of profits each year into an indi- viaul account for employees; some portion of which could fre- quently be withdrawn as income at the time the contribution was to be made; some portion of which could frequently be bor- rowed; and some portion of which would be paid to vested em- ployees upon termination of employment (retirement or other- wise) in a lump sum. -Thrift-savings and stock bonus defined contribution plans into which both employee and the employer would generally make contributions. The employee would generally have some access to at least a portion of the funds while employed and would receive a lump sum or a payout of stock upon job termination. The pre-ERISA pension plans gave the employee a benefit prom- ise with investment risk and reward inuring to the employer. The pre-ERISA capital accumulation plans gave the employee a benefit promise of whatever account balance existed at termination with investment risk and reward inuring to the employee. ERISA changed the definitions: * * * (2) The terms "employee pension benefit plan" and "pension plan" mean any plan, fund, or program which was heretofore or is hereafter established or maintained by an employer or by an employee organization, or by both, to the extent that by its express terms or as a result of surrounding circumstances such plan, fund, or program: (A) provides retirement income to employees, or (B) results in a deferral of income by employees for periods extending to the termination of covered employment or beyond.* * * ERISA, therefore, is viewed by some as having actually moved the Nation further away from a path of setting a national retire- ment income policy by including both plans which provide "retire- ment income to employees" and those which "result in the deferral of income" within the definition of "pension" plan. The sections that follow look at all of these plans over the span of 1975 to 1983 in the case of the number of plans, and 1972 to 1983 in the case of coverage by plans. Analysis is based upon the best available data from the Internal Revenue Service and the Bureau of the Census. Because annual data from the ERISA Form 5500 has not been published, the information is less complete than ERISA sought to make possible. First, the chapter sets forth conclusions regarding the impact of ERISA on plans and their attractiveness to employers and workers. This section looks at defined benefit plans, defined contribution plans, welfare plans, and noncovered plans. Second, the chapter looks at the 1974 to 1983 statistics on defined benefit and defined contribution plan creations and terminations. To the degree possible, the number of participants involved is as- sessed. However, available government data makes this analysis very difficult. Third, the chapter briefly explores ERISA and employer-spon- sored individual effort for capital accumulation through salary re- duction programs. Fourth, the chapter reviews the data that is available on cover- age by plans in 1972 and 1983 in an effort to see how changes in the plan universe have affected workers and likely benefit receipt. Finally, the chapter briefly reviews experience in 1975 and 1983 with the ERISA established individual retirement accounts and compares savings results of alternative "pensions." C. WHAT WAS THE IMPACT OF ERISA? ERISA required that all employers take pensions more seriously. It affected different types of plans in different ways. There is uni- versal agreement, for example, that it made defined contribution plans relatively more attractive than defined benefit plans from the employers perspective, as discussed below. From the workers perspective, ERISA made the defined benefit plan relatively more attractive because of the introduction of termination guarantees. Yet, the benefit security promised by ERISA has not been over- riding in determining employee choice. The analysis that follows must be read with factors that go beyond ERISA in mind.3 First, the period we are evaluating was the period during which the baby boom was entering the work force; younger workers change jobs frequently and prefer the faster vesting and "portabil- ity" of defined contribution plans. Second, by the time ERISA was passed most large employers al- ready had defined benefit plans and were beginning to establish de- fined contribution plans to complement them. Third, there have been two recessions since ERISA was passed and an ongoing employment shift from high pension sponsoring manufacturing industries to lower pension sponsoring service in- dustries. Fourth, these economic factors have had unions fighting just to keep what they have. Additionally, the period has seen a decline in union membership. Finally, international competition has caused employers to seek more cost containment in their employee benefit programs, leading many to prefer the fixed contribution approach. These factors make it impossible to say why certain trends have developed since the passage of ERISA. In spite of this, however, the affect of the law on the attractiveness of plan types can be explored and post-ERISA trends can be identified. The next two major sec- tions explore both issues.

1. ERISA AND DEFINED BENEFIT PLANS ERISA made it more likely that people covered by ongoing de- fined benefit plans would receive benefits and that they would un- derstand their plans: -Earlier participation, faster vesting, and joint and survivor benefit provisions increased participation and vesting. -Reporting and disclosure requirements assured the availability of information. -Minimum funding standards and the PBGC assured that most, if not all, benefits vested would be paid; and -Fiduciary standards sought to assure the security of assets once accumulated.

A full discussion of these factors can be found in the following publication. Salisbury, Dallas L. America in Transition: Implications for Employee Benefits. Washington, EBRI, 1982. Most benefits in society also have costs attached to them. ERISA made it more expensive to maintain a defined benefit plan: -Administrative costs were increased due to government forms, requirements for actuarial valuations, PBGC premiums, and disclosure to participants through actual document distribu- tion. -Plan maintenance costs were increased due to minimum par- ticipation and vesting standards which required that more funds be contributed or that all benefits be reduced. -Some plan sponsor financial flexibility was eliminated by mini- mum funding standards, PBGC termination liabilities, PBGC contingent liabilities, and restrictions on the use of plan assets; and -The Financial Accounting Standards Board (FASB) has used the liabilities "created" by ERISA as partial justification for requiring disclosure of pension plan assets and liabilities in corporate financial statements which has in turn affected cor- porate financial assessments. For both single employer and multiemployer defined benefit plans, surveys indicate that ERISA made both less attractive to employers. For the latter, the prospects for further growth were also affected by passage of the Multi-Employer Pension Plan Amendments Act of 1980 [MEPPAA] which significantly increased employer plan sponsor liabilities in the event that they entered an ongoing plan for the first time. On the other hand, MEPPAA is be- lieved to have stabilized plans in terms of employers already in them. It is significant to note that in spite of these changes defined ben- efit plans are still maintained and new ones are regularly estab- lished. Even though the data in later sections shows fluctuations, ERISA did not kill defined benefit pension plans. However, pending legislation and new proposals, such as those highlighted in other chapters in this paper, are likely to increase the costs of maintain- ing a defined benefit plan.

2. ERISA AND DEFINED CONTRIBUTION PLANS The minimum vesting and participation standards, reporting and disclosure requirements, and fiduciary standards of ERISA also apply to defined contribution plans. As a result, workers were given new legal protections and access to information in a common format. However, since ERISA respresented less of a change in estab- lished defined contribution plan practices, the imposition of mini- mum standards had far less cost attached to it for defined contribu- tion plans.

3. THE EMPLOYERS PRESPECTIVE Defined contribution plans were viewed by most plan sponsors in the pre-ERISA period as "savings" or "capital accumulation" vehi- cles, not the main source of retirement income to supplement Social Security. Most plans were designed to provide some access to the funds before retirement to meet special circumstances. As a result, participation and vesting standards were more "favorable" than was generally found in defined benefit plans. Therefore, they required less change to accommodate ERISA. Some corporations, it must be noted, have used profit sharing successfully as the retire- ment income plan, but most companies favored the defined benefit plan supplemented with "savings and capital accumulation" plans. As ERISA, as amended, increases the costs and burdens of de- fined benefit plans, surveys indicate that more and more compa- nies are becoming interested in either (1) moving to the defined contribution plan as primary, or (2) reducing what will be provided by the defined benefit plan in the future and placing new emphasis on the defined contribution plan. The fixed cost contribution concept has increasing appeal; as does the absence of any employer risk related to investment per- formance. The defined contribution plan may actually be more ex- pensive in the long run, but the cost is certain, will not be greatly affected by future legislation, and can be budgeted. Finally, issues related to unfunded liabilities and asset reversions become irrele- vant if all you have is a defined contribution plan.

4. THE WORKERS PERSPECTIVE ERISA also clearly established that for the worker interested in saving on a tax preferred basis, and interested in having access to all or part of the accumulated balance, the defined contribution plan was the only legal approach. Surveys of the population show that workers don't really begin to think about retirement income, as opposed to saving, until their midforties or early fifties. Since the baby-boom has been entering the work force during the post- ERISA period, this congressionally encouraged (intended or not) movement toward defined contribution plans has been welcomed by a large segment of the population. Defined contribution plans have a number of characteristics that appeal to those interested in seeing what they are getting and in- terested in having access to the funds in other than an annuity stream after retirement: there is a known account balance, an abil- ity to take it with you, and generally an ability to partially control investment decisions. Some DC plans, foremost among them the TIAA-CREF plan, don't allow this interim access and do pay benefits as an annuity. There are also profit sharing plans that fit this description. The vast majority of DC plans, however, have the "baby-boom at- tributes": -Defined contribution plans under ERISA can allow partici- pants to borrow funds from their account under certain cir- cumstances. -Defined contribution plans under ERISA can allow partici- pants to withdraw balances under certain legally defined "hardship" conditions. -Defined contribution plans under ERISA can allow partici- pants to take their entire balances upon preretirement termi- nation of employment and they may then treat the funds as taxable income or place them in another tax-prefered capital accumulation plan. -Defined contribution plans under ERISA allow participants to apply 10 year forward income tax averaging to the lump-sum distribution in order to reduce tax liability; and -Defined contribution plans under ERISA can allow partici- pants to direct their own investments.

5. ERISA AND WELFARE PLANs ERISA is a pension, health, and welfare plan statute. In spite of the fact that many provisions relate to health and welfare plans, however, the law has been implemented largely as a pension law. As a result, ERISA has had almost no affect on these other plans with the exception of requiring large welfare plans to meet some reporting and disclosure requirements. A few cases have arisen where the fiduciary standards of ERISA had to be used to protect participants interests.

6. ERISA AND NONCOVERED PLANS The period since 1974 has included an ongoing discussion of ex- tending ERISA requirements to plans now exempted from its provi- sions. This has been most intense with regard to public plans. And this discussion has had an affect: numerous political subdivisions have passed their own "reform" requirements in an effort to show the Congress that Federal standards are not needed. ERISA and public concern over pension benefit security have also led many political subdivisions to begin funding their plans on an ERISA equivalent basis, or to move in that direction (a study recently completed by the National Council on Teachers Retire- ment reviews actions taken in 40 States). Public and church units almost exclusively use defined benefit plans as their basic retirement income programs and then supple- ment with some deferred compensation-salary reduction arrange- ments. D. DEVELOPMENT OF EMPLOYER-SPONSORED RETIREMENT PLANS The main focus of ERISA was on defined benefit and defined con- tribution retirement income and capital accumulation programs: all termed pensions by ERISA. This section first reviews IRS data on (1) net plan creations-plan qualifications less plan termina- tions, (2) plan terminations, and (3) plan qualifications. Then it as- sesses the establishment of defined contribution plans by type in each of the post-ERISA years. It should be noted that the data tells us what has happened, not why it has happened. Further, due to the lack of ERISA Form 5500 data, it is not possible to assess actual participation levels by plan type. The next section does review coverage by all plans.

1. NET PLAN CREATION: 1974 TO 1983 The pre-ERISA universe of plans was 55 percent defined benefit and 45 percent defined contribution. By 1983, the relative positions had more than reversed. The universe of plans was 59 percent de- fined contribution and 41 percent defined benefit. During 1974, the last year during which ERISA provisions were not generally in force, 55 percent of net new plans were defined benefit, consistent with the makeup of the full universe of plans. During 1976, the first full year in which all ERISA provisions were generally in force, 100 percent of the net new plans established were defined contribution. The number of defined benefit plans in operation decreased in that year by 1.7 percent. During 1983, the latest full year for which data is available, 33 percent of net new plans established were defined benefit and 67 percent were defined contribution, indicating that the universe is continuing to shift in the defined contribution direction. Hopefully the government will complete current work to make ERISA Form 5500 data available so that the affect on participation levels can be evaluated (see table 1). TABLE 1.-NET PENSION PLAN CREATIONS AND GROWTH, 1974 TO 1983

Definedcontribution Definedbenefit Totalpension plans Year Annual Total Growthrate Annual Total Growthrate Annual Total Growthrate number created (percent)number created (percent)number created (percent)

1983...... 30,672 480,900 6.8 14,900 330,485 4.7 45,572 811,385 6.0 1982...... 47,054 450,228 11.7 23,146 315,585 7.9 70,200 765,813 10.1 1981...... 48,842 403,174 13.8 19,253 292,439 7.0 68,095 695,613 10.9 1980...... 41,511 354,332 13.3 14,552 273,186 5.6 56,063 627,518 9.8 1979...... 33,548 312,821 12.0 12,488 258,634 5.1 46,036 571,455 8.8 1978...... 45,295 279,273 19.4 5,103 246,146 2.1 50,398 525,419 10.6 1977...... 17,985 233,978 8.3 1,616 241,043 .7 19,601 475,021 4.3 1976...... 14,187 215,993 7.0 - 4,180 239,427 - 1.7 10,007 455,420 2.2 1975...... 11,162 201,806 5.9 10,769 243,607 4.6 21,931 445,413 5.2 1974...... 24,599 190,644 14.8 30,002 232,838 14.8 54,601 423,482 14.8 Source:Employee Benefit Research Institute [EBRI] tabulations basedon Internal Revenue Service data. The rate of new plan establishment has not been consistent during the post-ERISA period. As table 1 data shows, all net plan creation slowed during the years just following ERISA enactment. By 1978, defined contribution plan creation exceeded its pre-ERISA rate. Defined benefit plan creation has not reached its pre-ERISA rate but had climbed significantly by 1982. Plan creation and termination at least appears to be affected by legislative change, with net creation falling after each major change in the law which requires amendment of existing plans. For example, the Revenue Act of 1978 made "pension" changes, includ- ing the creation of section 401(k) salary reduction programs. The following years saw defined contribution plan formation slow. The act did not affect defined benefit requirements and plan creation continued to climb. The Tax Equity and Fiscal Responsibility Act of 1982 [TEFRA] made significant additional changes in ERISA ap- plicable to both types of plans. Growth of both slowed. While it is impossible to show cause and effect, the net growth rate for all plans was lower in 1983 than in any year since 1976 for defined contribution plans and since 1978 for defined benefit plans.

2. PLAN TERMINATIONS: 1974 TO 1983 The data on plan terminations since 1974 indicates that the pen- sion system is in constant fluctuation. The rate of terminations hit its highest level in the period just following passage of ERISA. From 1974 to the end of 1983, over 125,000 plans were terminat- ed (see table 2). Over 78,000 were defined contribution plans, as compared to a total of 191,000 of these plans in operation when ERISA was enacted. Just under 48,000 defined benefit plans were terminated during the post-ERISA period, against a base of 232,838 plans at year end 1974. Available data does not allow matching to see how old plans are when terminated or to establish if these ter- minations are pre-ERISA plans or turnover in post-ERISA plans. The series of Current Population Surveys show that throughout all this activity the total number of workers covered by plans has grown. This data is discussed in a later section of the chapter. TABLE 2.-PENSION PLAN TERMINATIONS, 1974 TO 1983

Definedcontribution Definedbenefit Totalpension plans Year AinnualTotal ended Annua Totalended Annual number number numberTotalended

1983...... 11,417 93,531 7,230 63,369 18,647 156,900 1982...... 10,108 82,114 5,043 56,139 15,151 138,253 1981...... 8,906 72,006 4,536 51,096 13,442 123,102 1980...... 8,982 63,100 4,297 46,560 13,279 109,660 1979...... 7,574 54,118 3,267 42,263 10,841 96,381 1978 ...... 10,661 46,544 4,625 38,996 15,286 85,540 1977...... 10,478 35,883 5,337 34,371 15,815 70,254 1976...... 6,943 25,405 8,970 29,034 15,913 54,439 1975...... 3,558 18,462 4,550 20,064 8,108 38,526 1974...... 2,207 14,904 2,577 15,514 4,784 30,418 Source:Employee Benefit Research Institute [EBRI] tabutations basedon Internal Revenue Service data. The rate of total plan terminations increased in 1983, with de- fined benefit terminations at their highest level since 1978, but still far below the 1974 to 1977 level. The rate for defined contribution plans was lower than in any other post-ERISA year. The volatility is far greater for defined contribution plans, which may be ex- plained by the fact that they are more readily sponsored by small businesses which have much higher business failure rates than large firms.

3. PENSION PLAN QUALIFICATIONS: 1974 To 1983 New pension plan qualification patterns have been consistent with net plan creation patterns. Defined benefit plan establishment has not reached the pre-ERISA level in either percentage growth or absolute terms in any later year (see table 3). Over 28,000 new plans were established in 1982, but the number dropped in 1983 to just over 22,000 plans.

TABLE 3.-PENSION PLAN QUALIFICATIONS 1974 TO 1983

Definedcontribution Definedbenefit Totalpension plans Year ______Annualnumber Totalcreated Annualnumber Totalcreated Annualnumber Totalcreated

1983...... 42,089 560,719 22,130 377,628 64,219 938,347 1982...... 57,162 518,630 28,189 355,498 85,351 874,128 1981...... 57,748 461,468 23,789 327,309 81,537 788,777 1980...... 50,493 403,720 18,849 303,520 69,342 707,240 1979...... 41,122 353,227 15,755 284,671 56,877 637,898 TABLE 3.-PENSION PLAN QUALIFICATIONS 1974 TO 1983-Continued

Totalpension plans Year Definedcontribution Definedbenefit Annualnumber Totalcreated Annualnumber Totalcreated Annualnumber Totalcreated

1978...... 55,956 312,105 9,728 268,916 65,684 581,021 1977...... 28,463 256,149 6,953 259,188 35,416 515,337 1976...... 21,130 227,686 4,790 252,235 25,920 479,921 1975...... 14,720 206,556 15,319 247,445 30,039 454,001 1974...... 26,806 191,836 32,579 232,126 59,385 423,962 Source:Employee Benefit Research Institute [EBRI] tabulations based on Internal Revenue Service data. The absolute number of new defined contribution plans estab- lished exceeded pre-ERISA levels in every year except 1975 and 1976. The growth rate, however, exceeded that for 1974 only in 1978. And the rate of plan establishment in 1983 dropped by over 35 percent from the previous year. Assessing these patterns relative to major legislation leads to the conclusion that change in the regulatory environment causes the pace of plan creation to slow. This may be because service provid- ers spend time amending plans during these periods rather than selling new ones, for example, in response to the passage of TEFRA. The numbers clearly show that the size of the pension uni- verse has been growing, in spite of changes in the regulatory envi- ronment. What we don't know is how the patterns would have been different in the absence of legislation. Additionally, we do not know how to factor in other demographic, economic and social changes. We do, however, have employer surveys which indicate that their behavior as plans sponsors is affected by changes, or the discussion of changes, in the legislative environment.

4. How BIG ARE NEW POST-ERISA PLANS? Annual surveys from the U.S. Bureau of Labor Statistics have in- dicated for most post-ERISA years that nearly all employers with more than 250 employees sponsor pension plans and that over 90 percent of those employed by these firms are covered by tax quali- fied retirement plans. For firms with fewer than 250 employees, coverage is well below *50 percent. As a result, this is where new plan formation has principally taken place. Table 4 shows average new post-ERISA plan size by plan type. Stock bonus plans and ESOP's were relatively larger than the aver- age new plan size. This is probably explained by the fact that each is a relatively new type of plan that is being used as a financing vehicle for many troubled businesses. Few employers or workers view them as retirement income programs. TABLE 4.-Average new post-ERISA plan size, 1976-83 Type of plan: Average size D efined contribution ...... 34 Stock bon us...... 1,201 E S O P ...... 2,085 P rofit sharing ...... 23 M oney purchase...... 12 D efin ed ben efit...... 76 Source: Employee Benefit Research Institute [EBRII tabulations of Internal Revenue Service data. The small size of both new defined benefit and defined contribu- tion plans indicates that small businesses are establishing pro- grams. Since small employers do, according to the Bureau of Labor Statistics, experience higher job turnover rates than large business- es, and more readily need to budget expenses, it is not surprising that they appear to more readily choose the defined contribution plan. These statistics do not provide information on why employers are doing what they are doing, or document the precise nature of the shift from one type of plan to another. The fact that employers are shifting is borne out by surveys of employers and by experience of the last 2 years with employers terminating defined benefit plans, recovering assets, and then establishing replacement defined con- tribution plans.

5. DEFINED CONTRIBUTION PLANS By TYPE: 1976 TO 1983 The data presented in the following sections highlight experience since 1974. The pattern of establishment of particular types of de- fined contribution plans represents evidence that the tax code itself can significantly affect both individual and corporate behavior. While this may appear obvious to some, it does not always get re- flected in policy assessment. For example, calculation of tax ex- penditures assumes that a change in the tax law will not change behavior. And many do seem to believe that favorable tax provi- sions for pensions could be drastically changed or eliminated with- out causing plans to disappear.

(A) EMPLOYEE STOCK OWNERSHIP PLANS AND STOCK BONUS PLANS Stock ownership plans were created pre-ERISA but their legal status was not entirely clear. As a result, fewer than 300 ESOP's were believed to be operating before 1975. ERISA clearly established the legal status of ESOP's and codified the stock bonus plan. Seven tax related acts of Congress have af- fected tax status since that time. Over 1,200 ESOP's, PAYSOP's, and TRASOP's in operation at the end of 1983 had been given IRS determination letters. The three different types of plans exist due to ongoing changes in the law regarding the nature of the favor- able tax treatment provided. Approximately 4,250 stock bonus plans that had requested IRS determination letters were in operation at the end of 1983; 330 plans were given determination letters in 1983, a growth rate of 6.9 percent. These plans are most frequently established as "leveraged ESOP's." The combined numbers shown in table 5 for the post-ERISA period indicate that the combination of clear legal standards and favorable tax status can make a difference.4

4 A detailed discussion appears in the following. Peckman, David A., Employee Stock Owner- ship Plans: A Decision Makers Guide. Washington, EBRI, 1983. 118 TABLE 5.-PLAN QUALIFICATIONS AND TERMINATIONS, 1976-83

Year ESOP TRASOP Stock Ended Net

1983 ...... 165 0 330 54 441 1982 ...... 132 0 251 42 341 1981...... 108 0 410 50 468 1980...... 0 51 482 37 496 1979 ...... 0 286 574 31 829 1978...... 0 196 850 42 1,004 1977 ...... 0 132 856 15 973 1976 ...... 80 85 758 0 843 Total ...... 405 750 4,511 271 5,395 Source:Employee Benefit Research Institute [EBRI] tabulations oftnternal Revenue Service data.

(B) PROFIT-SHARING PLANS Among the oldest and most widespread defined contribution plans are the profit sharing variety. Such plans have been in oper- ation for decades, with approximately 180,000 in operation at the time ERISA was passed. Profit-sharing plans are those into which an employer pays spe- cial current or deferred amounts based on business profits. ERISA defines all such plans as pension plans, but this is not strictly the case. Some of these programs act as sources to provide loans, to provide cash distributions during or upon termination of employ- ment, or to provide distributions at retirement. Table 6 presents IRS statistics for the post-ERISA period on the growth of profit-sharing plans. The universe of profit-sharing plans with IRS determination letters has grown by 89.8 percent in the post-ERISA period. This compares to 82.2 percent for all plans; 35.7 percent for defined benefit plans; and 138.3 percent for all defined contribution plans.

TABLE 6.-PROFIT-SHARING PLANS, 1976-83

Year Created Ended Netnew Total Gprcet

1983...... 25,251 6,727 18,524 341,711 5.7 1982...... 33,583 6,201 27,382 323,187 9.3 1981...... 31,827 5,430 26,397 295,805 9.8 1980...... 30,251 5,990 24,261 269,408 9.9 1979...... 24,239 5,107 19,132 245,147 8.5 1978 ...... 32,890 7,334 25556 226,015 12.7 1977...... 17.099 7,576 9,523 200,459 4.9 1976...... 11,129 193 10,936 190,936 6.1 Pre.1976...... NA NA NA 180,000 NA Total...... 206,269 44,558 161,711 341,711 NA Source:Employee Benefit Research Institute [EBRI] tabulations ofInternal Revenue Service data. The Revenue Act of 1978 extended the "salary reduction" provi- sions of the Internal Revenue Code (section 401k) (that have his- torically applied to profit-sharing plans) to allow "stand-alone" salary reduction plans. There is not yet good data on the number of these arrangements. This is because many employers are establish- ing these plans by amending existing thrift-savings plans rather than starting from scratch. A large number of employer surveys have been taken, however, which indicate that the rate of estab- lishment of these plans is proceeding at a very fast pace. Once Form 5500 data is available we will have a picture.

(C) MONEY PURCHASE PLANS These defined contribution plans are funded at a defined percent of compensation level for each employee. An individual account is maintained for each employee to which contributions are made. In- vestment experience, both gains and losses, cause the account bal- ance to change. Upon job termination or retirement, the employee is entitled to that portion of the account balance in which a vested nonforfeitable right exists. Data presented in table 7 shows that the universe of money pur- chase plans grew by 107.2 percent between 1975 and 1983: the high- est growth rate of any of the non-ESOP defined contribution ar- rangements. The growth rate in 1983, however, was at the lowest level since enactment of ERISA. This may be explained by the growing interest in employer sponsored IRA's and "salary reduc- tion" programs among small employers seeking a defined contribu- tion capital accumulation approach with minimum administrative requirements.

TABLE 7.-MONEY PURCHASE PLANS, 1976-83

Year Created Ended Netnew Total (Percent)Growth

1983 ...... 16,0 21 4,396 11,625 215,484 5.7 1982 ...... 22,800 3,727 19,073 203,859 10.3 1981 ...... 21,093 2,784 18,309 184,786 11.0 1980 ...... 19,706 2,955 16,751 166,477 11.2 1979...... 15,969 2,436 13,533 149,726 9.9 1978...... 21,821 3,276 18,545 136,193 15.8 1977...... 10,113 2,881 7,232 117,648 6.5 1976...... 6,206 60 6,146 110,416 5.9 Pre-1976...... NA NA NA 104,000 NA Total...... 133,729 22,515 111,214 215,484 NA Source,Employee Benefit Research Institute [EBRI] tabulations of Internal Revenue Service data.

E. EMPLOYER-SPONSORED INDIVIDUAL EFFORT ERISA did not provide a basis for employers to act as an agent or facilitator of individual savings on a tax favored basis. This ap- proach was introduced by the Revenue Act of 1978 for section 401(k) salary reduction programs and later for employer sponsored individual retirement accounts or simplified employer pensions (SEP's). As employers have become familiar with these approaches they have begun to embrace them. Employees may elect to defer a portion of their salary that would normally be paid in cash in a salary reduction plan. Section 401(k) of the Internal Revenue Code states that if such a plan meets a nondiscrimination test, employees will not be currently taxed simply because they had the right to take their compensation in cash; they would be taxed instead when the cash is taken out of the plan. As previously noted, data on the number of these plans in oper- ation will not be generally available from Internal Revenue Service data. This is because most are being established by amending exist- ing thrift-savings plans. On December 31, 1982, data from the EBRI/HHS May 1983 CPS 5 showed that just over 7 million work- ers had these plans available to them. Surveys by Hewitt Associ- ates and Peat, Marwick, Mitchell & Co. indicated that several hun- dred plans were in operation by the end of 1983 and that employers continue to put them in place at a fast rate. The ERISA required From 5500 does collect data on these plans, and by 1985 there will be good numbers available if the government processes the forms. F. DEVELOPMENT OF COVERAGE FOR EMPLOYER- SPONSORED RETIREMENT PLANS Employer-sponsored "pensions" are the most widespread retire- ment income programs complementing Social Security. Previous sections of this chapter have documented the growth in number of plans. This section documents coverage growth that has accompa- nied new plan formation. The most complete pre-ERISA data on private "pension" cover- age comes from the U.S. Department of Commerce Bureau of Census Current Population Survey taken in April of 1972. This survey was very similar to one funded by the Employee Benefit Re- search Institute and the U.S. Department of Health and Human Services and undertaken by the Bureau of the Census in May of 1983. These surveys show that coverage among full time private sector wage and salary workers increased from 23 million to 33.5 million. This segment of the work force grew from 48 million to 61 million during this 1972 to 1983 period. As a percent of all workers in this class the coverage rate increased from 47 to 55 percent. Among men the coverage rate went from 52 to 56.4 percent, while for women it rose from 36 to 52.4 percent. These surveys also allow a comparison of coverage for pre- and post-ERISA years by industry, employee tenure and employee age for men and women. Table 8 shows that coverage grew in every industry between 1972 and 1983. It is particularly significant that coverage for women in- creased dramatically, and moved them very close to male coverage rates in most industries.

TABLE 8.-COVERAGE BY INDUSTRY, PERCENT, 1972 AND 1983

1972 1983 Industry Total Male Female Total Male Female

Mining ...... 72 73 ...... 73 73 73 Construction...... 34 35 ...... 35 35 43 Manufacturing...... 61 66 47 71 72 67

5 The survey was designed and sponsored by the Employee Benefit Research Institute (EBRI) and the U.S. Department of Health and Human Services (HHS). It was conducted under con- tract by the U.S. Census Bureau in May 1983. EBRI Issue Brief No. 32 presents the first findings from the survey. 121 TABLE 8.-COVERAGE BY INDUSTRY, PERCENT, 1972 AND 1983-Continued

1972 1983 Industry Total Male Female Total Male Female

Transportation...... 45 46 34 57 59 51 Communications .. 82 83 79 87 89 84 Finance,Ins., Re ...... 52 60 45 66 63 68 Services...... 29 36 24 45 47 44 Trade-wholesale...... 48 51 33 55 57 48 Trade-retail...... 31 60 25 36 39 34

Seurce:1972 data frem "Coveraie and Vestini of Full-limeEmployees Under Private Retirement Plans", by Walter W. Kotodrobetz aodDonald M. LandayinSocial Security Bulletin, Nooember 1973 toe toll-time workers. 1983 data represents EBBI tabulations teem the EBRI/HH5 May 1983. CtPS PensionSupplement. Table 9 shows that coverage grew for every tenure group, with growth again being highest for women and moving them above male coverage rates of 1972, but not up to male coverage rates of 1983. TABLE 9.-COVERAGE BY TENURE, PERCENT, 1972 AND 1983

1972 1983 TENURE Total Male Female Total Male Femule

20 24 15 29 30 281...... 1 to4 ...... 40 45 33 49 48 50 5to9 ...... 57 62 46 65 65 65 0to146...... 64 67 56 73 75 68 158to19...... 71 74 64 78 81 73 20 plus...... 73 74 67 83 84 77 Source:1972 data from "Coverage and Vesting of Full-Time Employees Under Private Retirement Plans", by Walter W. KolodrubetzandDonald M. Landayin SocialSecurity Bulletin, November 1973, for full-time workers. 1983 data represents EBRI tabulations from the EBRI/HHSMay 1983, CPS PensionSupplement. Table 10 shows that coverage grew for every age group. Female coverage increased by approximately 25 percent in every age cate- gory. The most significant indicator from this table is the fact that for the youngest age groups female coverage rates are now equal to or exceed male rates. TABLE 10.-COVERAGE BY AGE, PERCENT, 1972 AND 1983

1972 1983 TgeNR Total Male Female Total Male Female

25...... 30 33 26 38 36 40 251to 29...... 47 51 37 53 51 56 30 to 34...... 51 55 39 58 59 56 3510to 39...... 53 58 38 62 64 60 401to 44...... 54 59 41 61 66 53 45 t 49 ...... 55 61 42 60 63 54 50t54...... 56 62 43 62 64 58 55t59...... 51 56 40 64 67 59 60 ...... 39 44 28 52 56 46 Source:1972 data from "Ceverage and Vesting of Full-timeEmployees Under Private Retirement Plans", by Walter W. Kolodrubetz andDonald M. Landayin SocialSecurity Bulletin, November 1973, for full-time workers. 1983 data represents EBRI tabulations from the EBRI/HHSMay 1983, CPS PensionSupplement. The post-ERISA growth in pensions has been significant in terms of the number of plans, the workers covered, and the level of vest-

37-569 0 - 84 - 9 122 ing. Table 11 shows that this coverage is distributed across all earnings groups, and that vesting rates follow a similar pattern. TABLE 11.-EMPLOYMENT COVERAGE AND VESTING: DISTRIBUTION BY EARNINGS FOR NONAGRICULTURAL WAGE AND SALARY WORKERS, MAY 1983

Numberof workers (thousands) I Earnings Employment Coverage Totvted

Total...... 80,289 47,372 27,603 $1to $4,999 ...... 10,014 2,433 358 $5,000 to $9,999...... 15,323 5,747 2,023 $10,000 to $14,999 ...... 17,827 10,328 5,484 $15,000 to $19,999 ...... 13,101 9,422 5,874 $20,000 to $24,999 ...... 10,283 8,159 5,641 $25,000 to $29,999 ...... 5,515 4,365 3,048 $30,000 to $50,000...... 6,611 5,547 4,072 $50,000 andover...... 1,615 1,317 1,106

Percentagedistribution across earnings groups Employment Coverage Totalvesting

Total...... 100.00 100.00 100.00 $1 to $4,999 ...... 12.47 5.14 1.30 $5,000 to $9,999 ...... 19.08 12.13 7.33 $10,000 to $14,999...... 22.20 21.80 19.87 $15,000 to $19,999...... 16.32 19.89 21.83 $20,000 to $24,999 ...... 12.81 17.22 20.43 $25,000 to $29,999 ...... 6.87 9.21 11.04 $30,000 to $50,000...... 8.23 11.71 14.75 $50,000 andover ...... 2.01 2.89 4.01 Excludesworkers without reported earnings. While pension coverage and vesting are a good measure of who will get something, the question at the heart of any evaulation of the pension system is whether or not people will receive benefits and in what form. Table 12 indicates that benefit receipt rates will be significantly higher for future retirees as a result of system growth and work force aging, than they are today or than they were pre-ERISA. TABLE 12.-ESTIMATED PERCENTAGE OF FAMILIES RECEIVING PENSION BENEFITS AT AGE 65, AND AVERAGE REAL BENEFITS, BY CURRENT AGE AND MARITAL STATUS

Allfamilies Marriedcouples Singlepersons Cohortage in 1979 Percentageto Average Percentageto Average Percentageto Average amountof amountot receiveren benefitget antiamount ot receivebenefit fit receivebenefit betit

25 to 34...... 71 $12,417 75 $14,541 65 $8,701 35 to 44...... 65 11,190 67 12,563 60 8,823 45 to 54...... 52 8,656 58 9,621 41 6,496 55 to 64...... 37 5,315 44 5,548 26 4,718 - Realdollars are calculated using1982 as the base year. Source:"Social Security: Perspectives on Preserving theSystem" (Washington, DC:Employee Benefit Research Institute, 1982), p. 90. 123 Since ERISA chose to define pensions as including defined bene- fit and defined contribution plans, it is also important to know how both types of plans provide income in the post-ERISA environment. Table 13 shows that a significant number of current workers cov- ered by employer plans in their current employment expect to re- ceive a lump-sum distribution upon termination of employment rather than an annuity. While the central concern of this paper is the private employer pension system that is covered by ERISA, the table also presents the data for public plans. This provides a basis for comparison, and a means to set a baseline of data in the event that regulation is extended to public plans as some have advocated.

TABLE 13.-TYPE OF BENEFIT DISTRIBUTION AND VESTING AMONG NONAGRICULTURAL WAGE AND SALARY WORKERS AND THE ERISA WORK FORCE BY AGE AND SECTOR, MAY 1983

Coveredworkers Totalpercent vested Percentexpecting Percentexpecting (thousands) ofcovered pension futurelump sum only ERISAwork force: Private sector...... 27,550 60.11 48.21 11.90 25 to 34 years...... 9,883 46.81 32.33 14.48 35 to44 years...... 7,779 63.07 51.38 11.69 45 to 54years...... 5,756 70.11 58.72 11.39 55to 59years...... 2,707 73.13 65.38 7.75 60 to 64 years...... 1,424 70.98 65.91 5.07 Public sector...... 10,507 84.90 64.20 20.69 25to 34 years...... 3,183 78.08 49.72 28.36 35 to 44 years...... 3,261 88.22 66.30 21.92 45 to 54 years...... 2,365 86.39 71.36 15.03 55 to 59 years...... 1,082 87.81 76.29 11.52 60 to 64years...... 616 91.70 79.29 12.40 Nonagricultural wage and salary workers: Private sector...... 36,458 51.02 40.32 10.69 Lessthan 25...... 5,246 19.63 11.87 7.77 25 to 34years...... 11,516 42.80 29.47 13.34 35 to 44years...... 8,576 59.77 48.61 11.16 45 to 54years...... 6,185 67.41 56.45 10.96 55to 59years...... 2,912 69.88 62.53 7.35 60 to 64 years...... 1,520 69.97 64.88 5.08 65 yearsand older...... 465 51.28 45.82 5.46 Public sector...... 13,072 77.33 57.50 19.45 Lessthan 25...... 1,092 45.67 28.25 17.43 25 to 34 years...... 3,723 73.66 46.56 27.10 35 to 44 years...... 3,656 82.89 61.50 21.39 45 to 54years...... 2,552 83.32 68.33 14.99 55 to 59 years...... 1,159 84.05 72.43 11.62 60 to 64 years...... 664 88.07 76.56 11.51 65 yearsand older...... 227 67.17 59.25 7.92 Source-Employee Beoefit Research Institute, preliminary tabulations of May 1913 EBRI/HHS CPS pension supplement. Some have questioned whether plans that offer lump-sum distri- butions should have been considered pension plans by ERISA. Do they provide retirement income is the question asked. Now, data is available. The EBRI/HHS May 1983 CPS found that 6.6 million workers reported having received a lump-sum distribution from a prior job. The data indicates that large numbers of employees do 124 take advantage of the "take the dollars with me" feature in these plans. Table 14 shows what those workers said they did with the lump-sum distributions. The data indicates that the distributions are most readily used for consumption, not for retirement income. While this data does not provide any more detail than is shown, it provides the first good indicator of actual behavior. TABLE 14.-Use of lump sum distributions,1982 Use: Percent Retirem ent...... 4.4 An n u ity ...... 0.8 In vested ...... 16.8 Bough t House ...... 10.1 Bough t C ar...... 4.9 Vacation ...... 3.2 Oth er use...... 63.4 Source: Employee Benefit Research Institute [EBRI] tabulations of the EBRI/HHS May 1983 CPS Pension Supplement. G. INDIVIDUAL RETIREMENT ACCOUNTS [IRA's] Individual retirement accounts (IRA's) were originally estab- lished in 1974 as a tax incentive for retirement savings for workers not covered by a pension plan. The Economic Recovery Tax Act (ERTA) of 1981 expanded IRA eligibility to all workers and in- creased the maximum allowable contribution. The EBRI/HHS May 1983 CPS also provides data on who takes advantage of IRAs, and how many there are.6 During 1975, the first year of limited IRA availability, Internal Revenue Service data show that 1.2 million of the approximately 48 million eligible individuals opened IRA's. In 1981, the last tax year IRA use was limited, 3.4 million tax returns included an IRA contribution. Universal availability in 1982 brought greatly expanded IRA growth and created a new and highly competitive market. IRS data for 1982 recorded 12.1 million contributions to IRA's in tax year 1982-more than three times the number contributed in 1981. The number of IRA's formed in the lowest and highest income categories were nearly the same for both 1981 and 1982. The pro- portion of the dollar value of contributions shifted towards those in the highest income bra-ket from 19 percent to 21 percent. Those with less than $20,000 in income deducted 19 percent of total con- tributions. This shift might be explained by the ERTA increase in contribution limits. Higher earners are more likely to be able to set aside the maximum amount. Studies of savings indicate increased savings as individuals grow older. Highest IRA utilization occurs between 55 and 59, while the greatest number of IRA's are held by those between 45 and 54. The median age for IRA's is approximately 46 compared to a working population median of approximately 36. In terms of the question of who will have more savings as a result of IRA's, the most relevant measure is the number who create them.

6 EBRI Issue Brief No.7 31 presents the first selected findings from the survey on IRA's and their relationship to other forms of retirement saving. Women at all earnings levels are more likely to establish IRA's than men. Among those who had worked with their current em- ployer between 3 and 5 years in 1982 and earn between $15,000 and $30,000, women were almost twice as likely as men to have estab- lished an IRA. Establishment of IRA's by those without employer pensions did increase between 1981 and 1982. IRS data report 3.4 million IRA's among those without employer pensions at the end of tax year 1981. The EBRI/HHS CPS reported 4.8 million IRA's among those not covered by employer pensions for tax year 1982. The DOL/HHS 1979 CPS reported 1.5 million at the end of the 1978 tax year. IRA utilization generally increases with individual earnings. IRA's and salary reduction programs-which allow an employee to take a reduction in salary which is then placed in a tax-exempt trust-represent savings in addition to a pension for the vast ma- jority of workers. More than 46 percent of those with IRA's have a vested pension right with their current employer (for those who are older this number is far higher). Rates of coverage and usage of these arrangements are lowest among those without employer pen- sion plan coverage: only 12.1 percent of those without pension cov- erage have established IRA's. The most recent "equity" debate surrounding IRA's relates to spousal IRA's-those which are available for nonworking spouses. Of the 16.7 million workers who reported having IRA's in the EBRI/HHS CPS, 2.99 million (17.9 percent) reported having a non- working spouse. Of those 1.72 million (57.5 percent) established spousal IRA's. Employer pension coverage appears to increase the likelihood of spousal IRA use. This is probably because those with pension cover- age are more likely to be full-time, full-year workers, of older ages, and therefore more able to save. The EBRI/HHS study shows that among covered workers establishing an IRA and eligible to open a spousal IRA, 61 percent do so. This compares to a 44 percent rate of spousal IRA use for those without employer pension coverage. Among those with a vested pension right, the use rate is 66 per- cent. H. HOW HAS PLAN GROWTH AFFECTED SAVINGS? Employer sponsored programs and individual effort are both greater today than they were at the time ERISA was passed. A natural question is whether or not it makes a difference in terms of retirement income delivery. The available data indicates that it does.7 Pension coverage constitutes the major source of savings for more than half of current pension participants. While 38.8 million persons, or 40.5 percent of the labor force, had little or no savings of their own in 1983, 18.2 million of them were covered by employ- er pensions. Since these persons had incomes just over half the size of those with some savings, pensions appear to distribute savings more equally across income groups than would be the case in their

' The following publication assesses these issues in greater detail, and provides information on the implications of changes in the tax law. Korczyk, Sophie M. Retirement Security and Tax Policy. Washsington, EBRI, 1984. absence. Table 15 shows that Federal tax law has been effective in encouraging retirement savings at lower income levels that could not otherwise be expected. TABLE 15.-SAVINGS, PENSION COVERAGE, AND INCOME, 1983

Covered- Notcovered Averageannual income Savingsstatus Number Percent Number Percent Amount Percent (millions) (millions) Pecn Auut Pret

Nosavings 3...... 18.2 19.0 20.6 21.5 $9,661 40.5 Somesavings...... 36.9 38.4 20.3 21.1 19,209 59.5 Total ...... 55.1 57.4 40.9 42.6 15,338 100.0 Personsare classified ashaving some savings or nosavings according to whether or notthey reported any asset income in thesurvey. Asset incomeincludes interest, dividends, rents, and r alties. Coveragerefers to employerplans only bot in thepublic and in theprivate sector and does not include holders of IRAand Keogh accounts. Includespersons reporting negative asset income. Source:Employee Benefit Research Institute calculations based on Bureauof the Census,"Current Population Survey" (May 1983). As compared to IRA's and salary reduction programs, pensions are widely based at all earnings levels and are more evenly distrib- uted across the earnings spectrum. This is true for both defined benefit and defined contribution employer-sponsored programs. Table 16 shows the relative likelihood of participation in differ- ent types of programs. It shows that employer sponsorship does in- crease the likelihood that an individual will have savings to com- plement Social Security. TABLE 16.-UTILIZATION RATES FOR VOLUNTARY CONTRIBUTION RETIREMENT PROGRAMS, EMPLOYER OR SELF-PROVIDED, ERISA WORK FORCE, MAY 1983 [In percent]

Own Emplor doesE ayer Employerpension plan Employerrovided I(a 41k) pln 401(k)

Total private...... 60.41 33.02 31.36 20.68 Participants...... :...... 63.34 33.81 32.42 22.84 Vestedworkers...... 68.64 37.46 37.78 27.24 Noncoveredworkers...... 54.77 37.05 28.56 18.76 Unionized...... 55.49 22.18 30.37 15.82 Nonunionized...... 61.98 36.64 31.41 22.04 Sizeof firm: Lessthan 25...... 62.43 46.38 33.84 22.18 25to 99...... 49.23 41.70 25.06 21.13 100to 499...... 63.59 35.98 28.91 20.73 500 to 999...... 58.93 31.26 27.53 21.56 1,000 ormore...... 61.85 31.49 34.14 21.38 Source:Employee Benefit Research Institute, preliminary tabulations ofMay 1983 EBRI/RIS CPSpension supplement.

I. CONCLUSION The ERISA pension system has grown significantly since passage of the ERISA. The patterns have been different; incentives have changed; and, benefit delivery has increased. In short, the affects of ERISA can be seen, even if they cannot be clearly separated from factors of demographics, economics, and attitudinal change. ERISA at 10. A time for the Nation to decide whether the pie can continue to grow or whether choices need to be made. A time to reaffirm the need for employer sponsored programs or to move more in the direction of individual effort and Social Security." A time to assess whether further change in minimum standards such as vesting and participation will increase benefit receipt or in- crease costs and discourage plan sponsorship. In short, a time of de- cision. ERISA established the ground rules. ERISA defined the "new" world of pensions. Some question the definitions, but they are the law. Should the law encourage "retirement income" pensions that will complement Social Security with annuity payments, or "defer- ral of income" pensions that may or may not provide for people over the length of their retirement years. This may be the central issue for policy debate before additional actions are taken in the name of "retirement income" that actually discourage "retirement income" pensions and encourage "income deferral" pensions. The data presented in this chapter clearly shows that legislative change does affect employer behavior; does affect patterns of plan termination and formation; and does affect decisions on what type of plan to emphasize. 9 Employer pensions are a "success" for tens of millions of people who now receive benefits from them or will in the future. They are not for those who do not have them. Great care must be taken in establishing the policy balance that will expand the former group, rather than unintentionally making it smaller.

; A history and analysis of the interrelationship of these programs is contained in the follow- ing. Schieber, Sylvester J. Social Security: Perspectives on Preserving the System. Washington, EBRI, 1983. A full explanation of the post-ERISA rules governing coordination of the programs is published in the following. Schulz, James H. and Thomas D. Leavitt. Pension Integration: Concepts, Issues and Proposals. Washington, EBRI, 1983. * A thorough analysis of the findings of the EBRI/HHS May 1983 CPS Pension Supplement is presented in the following, together with a policy chapter analyzing major policy proposals. An- drews, Emily. The Changing Profile of Pensions in America. Washington, EBRI, 1984. Chapter 6 CONTINUING POLICY ISSUES (Prepared by Donald S. Grubbs, Jr.,' F.S.A., George B. Buck Consulting Actuaries, Inc.) A. INTRODUCTION What are the continuing policy issues facing pension plans today? What are the possible solutions? This chapter will address these two questions. In 1974, some had the naive belief that ERISA would resolve all of the major pension policy questions. But ERISA itself recognized that this was not so. It directed the Secretary of Labor (section 513) to undertake research studies, including studies of the effects of ERISA "on the provisions and costs of pension plans," "the role of private pensions in meeting the economic security needs of the Nation," and "the operation of private pension plans including types and levels of benefits, degree of reciprocity or portability, and financial and actuarial characteristics and practices, and methods of encouraging the growth of the private pension system." The Sec- retary is required to submit an annual report to Congress on these studies, including "recommendations for further legislation." In addition, ERISA directed the establishment (section 3021) of a Joint Pension Task Force to study and review continuing pension issues. Some of the items that were recognized as continuing policy issues and specifically assigned to the Joint Pension Task Force in- cluded age discrimination related to vesting, portability of pension rights, plan termination insurance for small employers, and Feder- al preemption of State and local law. Unfortunately the Joint Pen- sion Task Force was never actually organized. The years of development of ERISA before 1974 had raised a va- riety of issues that the legislation did not significantly address, in- cluding coverage and portability. These issues are still with us. ERISA itself created policy issues. Some of its provisions have proven insufficient to solve the problems they were directed at. Other provisions created new problems not foreseen by Congress. Changes outside the private pension system have also raised issues for pension plans. High interest rates have been a major factor leading to excess assets available on plan termination. High rates of inflation have raised questions about adjusting pensions to keep pace. Amendments to the Social Security Act raise issues re- garding the coordination of pensions with Social Security.

I The opinions expressed herein do not purport to represent the author's employer or any or- ganization with which the author is affiliated. (128) Some fundamental issues will always exist, such as the proper balance between desired benefits and acceptable costs. Other issues rise because, with each new achievement, humans raise their aspi- rations. In addition to private sector pension issues, new questions have arisen about public employee plans at the Federal, State, and local levels. I do not herein address all of the issues, but I have tried to sum- marize the principal ones. While no one is capable of entirely es- caping his own viewpoints, I have endeavored to briefly summarize the principal differing views on all sides of the issues. Limitations of space have not permitted me to explore any of the issues in depth. B. COVERAGE

1. BACKGROUND There is no requirement that any employer establish a pension plan, or that a pension plan cover all employees of the employer. Lack of adequate coverage was recognized as a problem before ERISA. Articles appeared about it in 1970 and 1972. Congressman John Dent spoke about it. When asked about it in 1972, Senator Williams acknowledged the problem, but indicated that the first task was to solve the problems for those already covered. In 1981, the President's Commission on Pension Policy identified coverage as one of the key problems of the private pension move- ment. It reported that only 42 percent of employees were covered under a pension plan. Some of the uncovered were in jobs that would result in their becoming covered after attaining age 25 and completing a year of service under the 1,000-hour standard. Among employees who have satisfied the age 25 and 1 year of service re- quirement, they reported that 58 percent were covered by a plan and 42 percent uncovered. Some others report percentages some- what higher, but in any event a large portion of workers are not covered.

2. REMOVING DIsINCENTIVES TO COVERAGE Some believe that the enormous complexity of ERISA and the ensuing regulations and reporting requirements have created a dis- incentive for employers to establish or to maintain plans. While this has been reported to be the cause of a number of the many plan terminations which occurred shortly after ERISA, it is not clear how significant a factor this has been in deterring coverage. Some have proposed amendments in ERISA designed to reduce reg- ulatory burdens. Others argue that some of the proposed changes would reduce needed protections. Many believe that the plan termination insurance provisions of title IV of ERISA are a deterrent to coverage under both single em- ployer and multiemployer defined benefit plans. Some propose to modify or abolish title IV for this reason. This controversial area is discussed in a later section of this chapter. The maximum dollar limits on contributions and benefits under plans reduce the incentives of small employers to establish or maintain plans covering both their employees and themselves, and discourage their providing larger benefits for their employees. Some advocate removing the dollar limits, or at least simplifying the rules, to eliminate this disincentive. Others oppose any change because it would result in larger tax deductions.

3. ADDING INCENTIVES To ESTABLISH PLANS Some would increase the tax incentives for employers to estab- lish plans, by use of tax credits or other devices. Others oppose these approaches because of the loss of revenue.

4. EXTENDING COVERAGE UNDER EXISTING PLANS Pensions plans are now permitted to exclude employees under age 25 or with less than 1 year of service. Many plans use these exclusions, while other plans have no such restrictions. The pur- pose of such requirements is usually to reduce administrative ex- pense for groups of employees with high turnover. The Retirement Equity Act of 1984 (which may be enacted by the time this is printed) will lower the age 25 requirement to age 21, extending coverage under existing plans. Some advocate completely eliminating any use of a minimum age.

5. MANDATORY UNIVERSAL PENSION SYSTEM In 1970, an article by the author of this chapter, entitled "Pro- posal for Mandatory Pensions," advocated requiring all employers to cover every employee under a pension plan providing at least minimum benefits or, alternatively, to fund a second layer of bene- fits under Social Security for every employee. In 1981, the Presi- dent's Commission on Pension Policy recommended a very similar proposal, which it called "minimum universal pension systems" [MUPS]. MUPS would be a step forward in solving the coverage problem. Some opposed it as too modest, charging that it would provide only "peanuts pensions." Others opposed it as imposing too much cost upon employers. Proponents acknowleged that the additional pen- sions provided would be modest, and that it would create additional costs as large as 3 percent of payroll for some employers (smaller for most), but argued that the additional security provided for workers was worthwhile and was worth the additional cost. C. ACCELERATED VESTING REQUIREMENTS 1. BACKGROUND The words "vested" and "nonforfeitable" are generally used interchangeably. They describe a right to an accrued pension bene- fit that cannot be forfeited by termination of employment. ERISA required all plans to provide accrued benefits under one of the three alternative requirements: (1) 100 percent vesting after 10 years of service. (2) 25 percent vesting after 5 years of service, grading up to 100 percent after 15 years of service. (3) Graded vesting under the "rule of 45." Most plans have chosen 100 percent 10-year vesting, since graded vesting is more difficult to communicate and administer, particu- larly because of required "buy-back" provisions that apply to plans with graded vesting. TEFRA required "top-heavy" plans (most small plans) to provide either 100 percent vesting after 3 years of service, or 20 percent vesting after 2 years of service grading up to 100 percent after 6 years of service. Most participants in defined benefit pension plans are subject to the 10-year vesting requirements. Some plans have vesting earlier than required. Most defined contribution plans have earlier vest- ing, including some plans with 100 percent immediate vesting. 2. THE NEED FOR EARLIER VESTING REQUIREMENTS Many employees never stay with any employer for 10 years, as is now required for vesting under the plans covering most partici- pants. The loss of pensions is particularly acute for women, who commonly have shorter periods of service than men. The worker who changes jobs frequently will have as great a need for retirement income as the worker who stays with a single employer all of his working career. Because of lack of vesting, the former worker may receive no pension at all, or only a small pen- sion based upon service with the worker's last employer. Some have suggested that 100 percent vesting be required under all plans after 5 years of service, or after 3 years of service. Some propose immediate vesting, at least with respect to some minimum level of benefits, since this is the only way to protect all pension benefits earned. 3. COST OF EARLIER VESTING Earlier vesting increases the cost of pensions plans. At the re- quest of congressional committees, I conducted a detailed study of the cost of mandatory vesting provisions 2 being considered when ERISA was developed. These studies indicate that requiring full vesting after 5 years of service for plans already subject to ERISA's requirements would create increases in cost for most defined benefit pension plans ranging from zero to 7 percent of present plan costs and from zero to 0.2 percent of compensation of covered employees. Studies based on current data would probably produce similar re- sults. 3-year vesting would be more expensive than 5-year vesting. Any increase in cost is a negative consideration. But some argue that the cost increase is small and affordable, and that this disad- vantage is outweighed by the need for earlier vesting. Some argue that it is unfair to impose more rigorous require- ments and higher costs on employers that maintain pension plans, and greater cost increases on employers that voluntarily provide the most liberal benefits, since other employers provide no pen- sions at all. The MUPS proposal discussed above would require all

2 U.S. Congress. Senate. Committee on Labor and Human Resources. Subcommittee on Labor. Study of the cost of Mandatory Vesting Provisions Proposed for Private Pension Plans. U.S. Government Printing Office, 1973. employers to provide a minimum level of pensions, and would re- quire 100 percent immediate vesting for the required minimum pensions. Plans with benefits greater than the minimum could apply less rapid vesting to the excess. Thus the MUPS proposal would distribute required costs more uniformly among employers.

4. DE MINIMIS BENEFITS Some argue that vested benefits after very short periods of serv- ice are too small to be useful in meeting retirement needs, and that the cost of preserving and administering them is relatively large in proportion to the benefits. Small vested benefits are often paid off in a lump sum and spent, rather than being preserved to provide retirement income. To the extent this occurs, earlier vesting does not accomplish the goal of meeting retirement income needs better. This relates to questions of portability, discussed in the next section. Some advocate correct- ing this problem by requiring plans to preserve benefits for retire- ment needs rather than paying lump sums, but this would increase the administrative burdens on plan sponsors. D. PORTABILITY Portability is the ability to transfer an individual's vested pen- sion credits to a succeeding plan, IRA or central clearinghouse upon termination of employment.

1. THE NEED FOR PORTABILITY In the past, it was suggested that portability did not really matter if vesting provisions were adequate. Does it really matter whether a worker receives four small vested pensions from four employers instead of receiving one combined pension from a single source? There are several reasons why the lack of portability often causes the individual to receive less retirement income. First, there is a practical problem the employer faces in trying to administer small vested pensions for the terminated employee. The problems of keeping records of the small pension, of keeping track of the former employee over many years, of sending of terminated vested employee summary annual reports and notices about plan changes, of locating him when benefits become payable, and of ad- ministering the small amount when it becomes payable, make the administrative costs high in relation to the amount of the benefit. ERISA provides needed relief to the employer or other plan spon- sor, from the disproportionately high administrative costs for small vested pensions. However this relief works to the detriment of the terminating employee. Under ERISA a pension plan may "cash- out" the employee's vested benefits in a lump sum upon termina- tion of employment. Lump-sum distributions over $1,750 ($3,500 under pending legislation) can only be made with the employee's consent. However, employees are tempted to accept the amounts of- fered so that they may gain discretionary use over them. IRA's coupled .with the rollover provision provide portability for the employee who wants to use them, but most individuals receiv- ing lump-sum distributions spend them rather than rolling them over into an IRA for retirement income purpose. This is particular- ly true of lower income workers with the greatest need for a pen- sion in their later years. Even if the amounts are rolled over into an IRA, an individual can still draw upon the IRA prior to retire- ment, although there are tax penalties to discourage premature withdrawals. One approach to portability would be to require deposit into a rollover IRA and deny access until normal retirement age. Further, an annuity payout could be required. Another approach would pro- vide for a central pension fund into which these amounts could be channeled, preserving the pensions for the employee while still avoiding the administrative burden upon the employer. In the absence of mandatory rollover IRA's or a central portabil- ity fund, many employers will very naturally pay the vested amounts in a lump sum directly to the terminating employee. Second, in addition to this direct defeat of pensions, the lack of portability discourages employers from liberalizing their vesting provisions beyond ERISA's minimum requirements. An employer whose objective is to provide a retirement income for employees has no incentive to move to earlier vesting if the result is small pensions vested after a few short years, which must be cashed out to solve the practical problems of administering small vested pen- sions. Such earlier vesting does not provide the retirement income the employer intends. If mandatory rollover IRA's or a central clearinghouse were established to accept and administer small pen- sion amounts, it might make earlier vesting feasible from a cost- benefit standpoint for additional employers. Mobile workers would have a greater opportunity to receive retirement income reflective of most of their years of employment rather than just their longest job. 2. POTENTIAL PROBLEMS OF A PORTABLE PENSION SYSTEM (A) ADMINISTRATIVE BURDEN UPON PENSION PLANS If a portable pension system were so designed or operated that it significantly increased the present administrative burden upon plan sponsors and administrators, this could be expected to in- crease plan terminations and decrease the formation of new plans. This would be most damaging to workers and to the private pen- sion movement. (B) ADMINISTRATIVE COST TO THE GOVERNMENT Any new program has costs for the Government to operate it. However, the difference in cost of operation between a program carefully designed for efficient administration, utilizing appropriate existing facilities, and one not so designed, can be very significant. This is important, regardless of whether the cost is paid by the Government or passed on to participants by the system. (C) CASH FLOW PROBLEMS FOR PLANS Most defined benefit pension plans normally provide their bene- fit payments in the form of a monthly income payable for life, and not in the form of lump-sum payments. Generally, only small amounts, previously referred to, are made available as lump-sum payments. Therefore, the flow of benefits from the plan if fairly regular, usually not exceeding the current year's contributions and investment income. Thus the plan usually has a positive cash flow and there is no need to liquidate investments to make benefit pay- ments. If a portable pension system were to require plans to make large lump-sum transfers to rollover IRA's or a portable pension clear- inghouse in lieu of making monthly pension payments, this could force plans to liquidate investments to make the transfers. Such transfers could adversely affect the funded status of benefits re- maining in the private pension plan, particularly if it were neces- sary to liquidate investments at a time when market values were temporarily depressed. This could reduce the benefit security of the remaining partici- pants, and could increase the potential liability of the Pension Ben- efit Guaranty Corporation. 3. REQUIREMENTS FOR MAXIMUM SUCCESS To be most successful a Federal portable pension system should have the following characteristics: (1) Utilization of amounts otherwise payable in a lump sum to provide a life income instead. (2) Assurance that the expected pensions will be paid. (3) Maximum pension income for participants from the amount transferred to the fund. (4) Pensions which meet the problem of inflation. (5) Equity to both employers and employees. (6) Capacity to be easily understood by employers and employees. (7) Minimum administrative costs for the portable pension system. (8) No significant administrative work or cost or other adverse consequences for private pension funds; and (9) High acceptability.

4. ISSUES IN DESIGN OF A PORTABLE PENSION SYSTEM Many issues are involved in the design of a portable pension clearinghouse, including the following: (1) Should use of the system be mandatory or voluntary on the part of plan sponsors and of individual participants? (2) What amount should be transferred from the employer pen- sion plan? (3) Should the portable pension fund operate as a defined benefit plan, a defined contribution plan, or both? (4) When should benefits be paid from the fund? (5) What forms of benefits should be provided, and how should the form be elected? (6) What annuity factors should be used to convert account bal- ances into monthly income? (7) How, if at all, should benefits be adjusted for inflation? (8) What death benefits should be paid upon death before retire- ment? (9) How should beneficiaries be determined? (10) How should fund assets be invested? (11) Who should administer the system, and how? (12) Who should pay administrative expenses of the system? (13) What modifications should be made in ERISA provisions con- cerning repayment of distributions upon reemployment? (14) How should benefit payments be taxed? The use of mandato- ry rollover IRA's for a portable pension system would raise most of these same questions. This chapter is too brief to explore these important questions of design of the system. The author has previously prepared a paper analyzing these questions, alternative solutions and considerations.

5. ENCOURAGING VOLUNTARY USE OF IRA ROLLOVERS Greater voluntary use of IRA rollovers would also extend porta- bility. Possible ways of doing this can be explored. Present laws regarding taxation of benefits may tend to discour- age rollovers. On the one hand, rolling over a pension plan distri- bution into an IRA may defer receipt of the money until a year when the individual is in a lower tax bracket; this encourages roll- overs. On the other hand, distributions from IRA's are all taxed as ordinary income, while lump-sum distributions from pension plans are taxed at lower rates under 10-year-forward-averaging rules. Some urge elimination of 10-year-forward-averaging in order to reduce the incentive against rollovers. Some also advocate this change in the interest of tax equity or reduction of revenue loss. Others would oppose it on the grounds that paying out in a single year the accumulation of many years pushes the taxpayer into a higher tax bracket, requiring special relief. Since this change would increase taxes for some, vigorous opposition would be expect- ed, particularly if relief is not provided by transition rules. Another deterrent to rollovers is the prohibition against rolling over the portion of a lump-sum distribution representing a return of the employee's own contributions. Some advocate removing this prohibition in order to increase portability. IRS representatives op- posed this in the past, on the grounds that it would make it more difficult to determine the taxable portion of the distribution. Advo- cates of the change argue that the determination is a simple one already commonly used for annuities, and that the convenience of IRS is not a suitable basis for determining social policy. E. PENSION ACCRUALS AFTER NORMAL RETIREMENT AGE

1. PRESENT PRACTICE UNDER DEFINED CONTRIBUTION PLAN Under defined contribution plans contributions are usually made for participants every plan year, although under a profit-sharing plan there may be years when no contribution is made. If contribu- tions are made during a year, they may be made for all partici- pants, or for all participants in a particular class. Under such plans, contributions are generally made without respect to age. Under some plans, however, no contribution is made for partici- pants older than normal retirement age, which is usually age 65. 2. PRESENT PRACTICE UNDER DEFINED BENEFIT PLANS Under defined benefit plans, the accrued benefit is defined as an amount of pension payable beginning at normal retirement age, which is usually age 65. Benefits are commonly defined as a benefit per year of service, such as 1 percent of average salary multiplied by years of service. Many plans have a maximum number of years of service counted for determining benefits, such as a maximum of 30 years. Other plans define the amount of benefit payable at normal retirement age, but reduce it for short service employees, such as a benefit of 50 percent of average pay reduced 1/25th for each year less than 25. Defined benefit plans differ in how they treat employees who work past normal retirement age. There are three common ap- proaches used. Many plans freeze the amount of pension payable at normal re- tirement age. Upon actual retirement later, the participant re- ceives the same monthly pension he would have received if he had retired at the normal retirement age. The additional years of serv- ice and compensation earned do not increase his monthly pension. This approach is taken by plan sponsors who want to encourage employees to retire at normal retirement age rather than continue working later, for a variety of personnel policy reasons. A second group of plans calculate the benefit upon later retire- ment taking account of the additional years of service and compen- sation earned. Under this approach, if two employees are hired today at the same salary at age 45 and 50, and if both work exactly 20 years at equal salaries and then retire at ages 65 and 70, respec- tively, both will receive the same monthly pension. This appears fairer than the previous method, which would provide a smaller pension to the employee retiring at age 70. A third group of plans endeavors to make the cost of a partici- pant's benefit the same as if he had retired at normal retirement age. Because he retires later and will not receive his pension for as many years, the plan can afford to pay him a larger pension. The amount of monthly pension that would have been payable at normal retirement age is increased by an actuarial factor, compa- rable to the actuarial reduction often used for early retirement. This method has often been used under plans funded with insur- ance or annuity contracts, where the reserve accumulated at normal retirement age is applied to provide the benefit. The in- creases for deferred retirement under this method are sometimes larger, and sometimes smaller, than under the second method de- scribed above. Under all three methods the pension plan costs are usually re- duced by an employee's decision to defer retirement, with the larg- est cost reductions occurring under plans that freeze the amount of benefit at normal retirement age.

3. DEFINITION OF NORMAL RETIREMENT AGE The large majority of pension plans have long set the normal re- tirement age as 65, sometimes coupled with a minimum of 5 or 10 years of service. Age 65 was selected primarily because this was the normal retirement age under Social Security, and private employer pension plans are usually intended to supplement benefits from Social Security. Under ERISA, a plan's normal retirement age may not be later than the later of attainment age 65 and completion of 10 years of participation in the plan. In 1983, the retirement age under Social Security was increased from 65 to 67, phased in over a period of years. This raises the issue of whether ERISA should be amended to allow private pen- sion plans to have the same normal retirement age as Social Secu- rity. Some argue that it makes sense to have benefits begin at the same time under both Social Security and private pension plans. People are living longer, and are in better health at older ages. Congress had determined that age 67 is a suitable normal retire- ment age for the Nation. Some would oppose this as a reduction in benefits compared to the present standard of age 65. ERISA forbids any cutback related to benefits already accrued, but not reductions in future accruals. Increasing the normal retirement age would substantially reduce the cost of providing pensions. It would be possible to apply this cost reduction toward providing postretirement increases to amelio- rate the impact of inflation. One proposal would allow plans to use a normal retirement age of 67 only if the plan provides automatic annual adjustments after age 67 of the lesser of the annual CPI in- crease or 3.6 percent. Others think age 67 normal retirement age should be allowed without restriction. If plans are allowed to use a normal retirement age of 67, one question is whether to require it to be phased in the same as Social Security, or for simplicity to allow an immediate increase to 67 for all, but with protection of benefits already accrued.

4. AGE DISCRIMINATION IN EMPLOYMENT AcT The Age Discrimination in Employment Act (ADEA), covering employees between ages 40 and 70, bans mandatory retirement and outlaws age discrimination in hiring, job retention, compensation, and conditions and privileges of employment. There are proposals to extend ADEA to ages above 70, forbidding the present practice of some employers of having mandatory retire- ment at age 70. Advocates of the change want to end age discrimi- nation. Opponents argue that it is very difficult for an employer to discharge an older worker who is no longer performing satisfactori- ly unless there is a mandatory retirement age. A change in the Federal law would affect relatively few employees, since few em- ployees now continue working to age 70, since mandatory retire- ment at any age has already been outlawed by a growing number of States, and since many employers already allow employees to work past 70 in other States. ADEA states that it generally is not unlawful to observe the terms of a pension plan. Department of Labor regulations under ADEA permit the present practice under defined benefit plans of freezing the benefit accruals at normal retirement age. A defined contribution plan may exclude an employee over normal retire- ment age from receiving employer contributions unless the plan is

37-569 0 - 84 - 10 a "supplemental plan," i.e., a plan which supplements benefits pro- vided by another plan. The EEOC has announced it will revise the regulations to require accruals to age 70. 5. REQUIRING PENSION ACCRUALs AFTER RETIREMENT AGE Some advocate requiring pension plans to accrue benefits after normal retirement on the same basis that applies to those under normal retirement age. The principal argument is fairness. In most plans, this would have a minor cost increase, but the in- creased plan cost would still usually be less than if employees actu- ally retired on their normal retirement dates. However, some em- ployers prefer to discourage deferred retirement for various person- nel reasons, such as creating promotion opportunities for younger workers and replacing older workers with younger workers. Young- er workers replacing the retirees may have lower salaries, as well as lower costs for group insurance. Some suggest that benefit accrual after normal retirement age should not be required for those plans which provide actuarial in- creases after normal retirement age, since this latter approach is generally as liberal as the former. A related question is whether defined contribution plans should continue to be allowed to exclude from participation employees first hired after normal retirement age. A similar question exists for defined benefit plans, which now may exclude employees hired within 5 years of normal retirement age. This provision reflects the fact that pensions are generally more expensive for older workers, and the related concern that re- quiring their coverage would discourage employers from hiring older workers. But some argue that the ability to require a mini- mum period of 10 years of participation solves the cost problem, and no exclusion from participation is needed. F. BENEFITS FOR SPOUSES AND SURVIVORS 1. FAMILY AND WORK PATTERNS On average, women workers are paid less than men. Their work- ing careers are often broken by a period of full-time homemaking and raising children. A larger proportion of women than men work in jobs not covered by pension plans. For those reasons, the pen- sions women earn on the basis of their own work histories general- ly tend to be smaller than those of men. Formerly it was commonplace for the husband to be the bread- winner, with the wife working at home her entire lifetime. This lifestyle still exists, but now represents a minority. In many fami- lies, both husband and wife have paid employment, but with the wife withdrawing from the paid workforce for a period of years to raise children. Many marriages end in divorce, with or without a later remarriage. A significant minority of workers never marry. Some people live together in relations similar to marriage but are unmarried. Many people are mistaken about whether or not they are legally married because they do not know the applicable State law. Some States have community property laws and others do not. Within this diversity, efforts to develop pension laws which are equitable and which meet the needs of people are difficult indeed. 2. PENSION PLAN BENEFITS FOR SPOUSES AND SURVIVORS Before ERISA, most pension plans defined the amount of retire- ment income payable to a worker, with no specific benefit for a spouse. Most plans provided a contingent annuitant option (often called a joint and survivor option), under which a worker retiring at normal retirement age or early retirement age could elect to re- ceive a reduced pension, with part or all of the reduced pension continued to a contingent annuitant (usually a wife) for life if the retired worker died first. In some of these the pension was payable jointly, requiring joint endorsement, but in most cases payment was solely to the retiree until the retiree's death. ERISA required most defined benefit plans and some defined contribution plans to offer a joint and survivor annuity option. If the retiring employee makes no election, ERISA also requires these plans to assume that the retiree elected the joint and survivor option with the spouse as contingent annuitant; this latter require- ment ordinarily has no effect whatsoever, since most plans require the retiree to actually make an election before payments com- mence. The Retirement Equity Act of 1984, not enacted at the time of this writing, will expand slightly the plans required to provide a qualified joint and survivor benefit. The Retirement Equity Act of 1984 will require the plans to pro- vide a joint and survivor annuity unless both the employee and the employee's spouse elect not to receive the joint and survivor bene- fit, or unless the employee is unmarried, unable to locate the spouse or has other valid reasons the spouse cannot sign. Neither ERISA nor the Retirement Equity Act require a joint and survivor annuity to be paid jointly (requiring both signatures), or for the spouse to receive any portion of the annuity while the retiree is living. Many employers and many multiemployer welfare plans provide death benefits outside the pension plan. These benefits are funded through life insurance or trust funds. They take the form of a lump-sum death benefit, often equal to one or two times annual earnings, or of a survivor annuity payable to the spouse and/or children of the worker. Such benefits outside the pension plan are not affected by ERISA's joint and survivor provisions, and are generally not coQ- sidered in this chapter. Such benefits are an important source of survivor protection, however. Any survey of the extent to which the needs of survivors are met must consider benefits from all sources. Before ERISA, many pension plans provided benefits to benefici- aries upon the death of the participant. Under one type of provi- sion, if a participant died while eligible for early retirement, he was assumed to have retired immediately prior to his death and to have elected the 50 percent joint and survivor option. The purpose was to treat the employee as though he had foreseen his death and made an intelligent choice among the benefits he was then entitled to, rather than forfeit his benefits because of a sudden and unfore- seeable death or because he didn't understand his choices. An illustration will assist in understanding the benefit. John Doe, hired at age 25, is covered under a pension plan providing a monthly benefit upon retirement at age 65 of $10 times his years of service. If he works 40 years to age 65 his monthly pension will be $400 (40 x $10). If he terminates employment after 30 years at age 55, he has accrued a monthly pension of $300 (30 X $10) payable be- ginning at age 65. If he elects to have his benefit begin at age 55, it will be actuarially reduced to have the same cost as the $300 begin- ning at age 65, reflecting the fact that payments begin sooner and continue longer. This actuarial reduction for early retirement re- duces the monthly amount from $300 to approximately $150. If John Doe elects a 50 percent joint and survivor annuity instead of a single life annuity, his monthly pension is further reduced to ap- proximately $140 monthly, to reflect the cost of the benefit to the survivor. If John Doe dies after his $140 monthly pension begins under the 50 percent joint and survivor option, his wife will receive 50 percent of his monthly pension, or $70 monthly. Under the above pension plan, if John Doe dies at age 55 before retirement, the plan assumes that this has taken place and pays his widow $70 montly for life. The benefit to the spouse would be larger if John Doe had continued working and died closer to age 65. If he dies at age 65 before retiring, for example, his widow would receive $180 monthly instead of $70. ERISA required almost all defined benefit plans to include a pro- vision similar to this. It is commonly called the preretirement sur- vivor annuity benefit. ERISA provides two options concerning how the cost of the provision will be borne. Under one option, which has been used by most plans, the cost is borne by the plan sponsor, as it was under the pre-ERISA provision of this type. Under the other alternative, participants may elect to be covered by this provision during the period they are eligible for early retirement; those elect- ing the coverage pay its cost by agreeing to have their benefit upon actual retirement reduced. Many plans that use this latter ap- proach reduce pensions 0.05 percent for each year of coverage, e.g., a 5 percent reduction in the pension payable at 65 if the partici- pant was covered during the 10 years from ages 55 to 65. The Retirement Equity Act will extend the period of survivor an- nuity protection to vested participants not yet eligible for early re- tirement.

3. PAYMENTS TO DIVORCED OR SEPARATED SPOUSES ERISA forbids the assignment or alienation of benefits. Creditors may not obtain a court order attacking a participant's pension. In divorce and separation proceedings, local courts have often ordered pension plans to pay benefit to a divorced or separated spouse of a participant. Some orders have directed payments in a manner or at a time in which no benefits could be paid to the participant him- self. Such orders placed plan administrators in the dilemma of dis- obeying either the local court or ERISA's nonalienation require- ment. Some plan administrators took each such dilemma to the U.S. District Court to resolve on a case-by-case basis. In an effort to resolve this problem, the Retirement Equity Act of 1984 defined a "qualified domestic relations order" as the kind, and the only kind, of court order a plan is directed to pay. In the absence of a valid court order, a spouse has no legal right to any benefit otherwise payable to a participant.

4. THE COMPLEXITY ISSUE The laws and regulations concerning the joint and survivor bene- fits are quite complex. This is particularly true of the notice and election procedures, and the new qualified domestic relations order provisions. This complexity reduces understanding of the require- ments and increases the difficulty and cost of communicating and administering the plans. Some advocate changes to simplify the area without reducing substantial benefits. Opponents fear that any change may reduce substantive rights.

5. CosT OF PRERETIREMENT SURVIVOR ANNUITY Some advocate requiring the preretirement survivor annuity to be provided automatically without any reduction in participant benefits to reflect the cost, at least for those eligible for early re- tirement. They argue that the benefit for those already eligible for early retirement only provides the benefit which they were already entitled to elect to receive. In fact even the participant who does not elect the preretirement survivor annuity coverage can still elect to actually retire early under the joint and survivor option, protecting his spouse. Thus there is no additional cost for the pre- retirement survivor annuity benefit, except for the individual who dies unexpectedly or without an understanding of the plan, who should not be deprived of the same benefit available to those who foresee death and make intelligent elections. The present system leaves some participants uncovered by the protection because of unwise elections concerning a matter that is difficult to under- stand. Opponents may argue that the survivor benefits do in fact have a cost, and that the change would be an increase in cost for employ- ees.

6. COMMUNITY PROPERTY CONCEPT Some advocate applying the community property concept to all pensions. Under this concept, 50 percent of the pension earned by any participant while the participant is married would be the prop- erty of the participant's husband or wife. The ultimate application of this would be to require half of a married participant's pension to be paid to the participant's hus- band or wife, even if living together, or perhaps that the pension be paid jointly, requiring endorsement by both, if they elect. Most advocates of the concept would make a more limited application, applying it only to divorced or separated spouses. Working out such a concept would require resolving the plan's lack of data on marital status, including the dates marriages began and ended. This-is complicated by the fact that individuals them- selves may misunderstand their legal status. For example, some 142 have married less than 3 days after obtaining the license in States where that is not permitted, or too soon after a divorce in States with limitations. Questions about the validity of foreign divorces, and the diversity of State laws regarding common law marriage, add to the confusion. A study of the problems occurring in probate is indicative of the difficulties. This does not mean that the difficul- ties can not be solved, but that the area would need careful study by experts in family law, with specific solutions proposed and an analysis of their effects. Even if the plan administrator had exact and reliable informa- tion, rules would be required to determine the portion of the ac- cured benefit to be assigned to the period of marriage. A variety of theories are available for this process, but some would involve data problems and substantial administrative work. The simplest solu- tion would be to allocate the total accrued benefit in proportion to the period of marriage. Some oppose this approach on the grounds that a court is a better place to determine the disposition of a couple's assets. The court addresses the division of pension benefits as part of the total problem of dividing property, taking account of the needs of the parties and other considerations. But the proponents point to ex- amples where the court system has worked poorly. G. INTEGRATION OF PRIVATE PENSION PLANS WITH SOCIAL SECURITY

1. THE CONCEPT OF INTEGRATION If the benefits or contributions of a plan are coordinated with those under Social Security, the plan is said to be "integrated." An employer may have an objective of providing pension benefits which will enable long service employees to retire on a standard of living approaching that which they enjoyed prior to retirement. For example, an employer may want to provide benefits which, to- gether with Social Security, will replace 70 percent of the employ- ee's preretirement earnings. If the average employee receives a Social Security benefit of 42 percent of pay, the employer might also provide a pension of 28 percent of pay, making a total pension of 70 percent of pay. But Social Security provides far less than 42 percent of pay for employees earning more than the taxable wage base. The Social Security benefit for an employee earning $50,000 per year is only about 16 percent of pay. Thus if the employer pro- vided all employees a pension of 28 percent, the employee earning $50,000 per year would have a total income from both sources of only 44 percent of pay, far less than the 70 percent objective. The employer's cost should also be considered. In 1984, emloyers pay Social Security tax of 5.4 percent (excluding the 1.3 percent for hospital insurance) of the first $37,800 of earnings, and zero of any excess. If an employer wants to spend an equal percentage of pay for each employee for his combined Social Security and private pension, it can afford to spend more for private pensions related to pay in excess of the taxable wage base. The above discussion assumes that the employer's objective is to provide a benefit program to meet the needs of employees, which also furthers employer objectives of attracting and retaining em- ployees. But many smaller businesses establish pension plans pri- marily for the purpose of deferring taxation for the owner or owners and other key employees, while desiring to minimize plan benefits and costs for lower-paid employees. These employers use integration as a tool to help accomplish this end. There are three general methods to integrate private pension plans with Social Security. First, the amount of contribution to a defined contribution plan can be integrated with the Social Securi- ty tax. For example, the employer might contribute 10 percent of total pay less the amount of Social Security tax paid, e.g., in 1984 the employer would contribute 4.6 percent (10 percent less 5.4 per- cent) of the first $37,800 of pay and 10 percent of any excess. The second method of integration, used for defined benefit plans, is to first calculate the pension under the plan's benefit formula and then reduce or "offset" this amount by some percentage of the benefits payable under Social Security. For example, an offset plan might provide pensions of 50 percent of pay less 50 percent of the primary insurance amount payable under Social Security. The third method of integration is the "excess" plan. A pure excess plan provides pensions based on pay above a certain level, usually the level upon which Social Security benefits are consid- ered to be based. IRS tables, recognizing that Social Security bene- fits are based on average earnings, including years when the tax- able wage base was much lower than today, assume that employer taxes provide Social Security benefits equal to 37.5 percent of the first $12,600 of pay for those retiring at age 65 in 1984. Using this assumption (which is not precisely correct), a private pension plan may provide 37.5 percent of the pay in excess of $12,600. More common than the pure excess plan is the "step-rate" plan, which provides one rate of pensions based on pay up to the integration level and a higher rate of pensions based on pay over the integra- tion level. For example, the plan might provide pensions of 40 per- cent of the first $12,600 of earnings plus 77.5 percent of earnings in excess of $12,600.

2. INTEGRATION AND THE INTERNAL REVENUE CODE The Internal Revenue Code provides that contributions or bene- fits provided under a plan may not discriminate in favor of employ- ees who are officers, shareholders, or highly compensated. A plan may pay larger pensions (measured in dollars) for highly compen- sated employees than for others, but it may not provide a larger percentage of pay for highly compensated employees than for others. A plan is not considered discriminatory merely because it excludes employees earning less than the Social Security taxable wage base, or merely because the contributions or benefits based upon earnings in excess of the taxable wage base differ from contri- butions or benefits based on earnings not exceeding the taxable wage base. Thus the code permits integration of contributions or benefits with Social Security. But the combination of benefits under the em- ployer's plan plus the portion of Social Security benefits funded by employer taxes may not discriminate in favor of the highly paid. The IRS has published very extensive and complex rules to deter- mine when this general concept is satisfied.

3. POSSIBLE CHANGES IN INTEGRATION REQUIREMENTS The IRS rules have been widely criticized as outdated, unfair, and far too complex. There have been many proposals to change the rules. Every one agrees that the rules are outdated in some respects. They fail to recognize changes in Social Security benefit levels that have occurred since 1971. If no other changes are made, it can be argued that updating to reflect current law is in order. But any change in the rules might require amendment of hundreds of thou- sands of plans, and it is not clear whether the differences would be sufficient to justify the expense. Some want to relieve inequities. There are disagreements about what is or is not inequitable. But all agree that the failure to at- tribute part of the Social Security benefits to any previous employ- er is fundamentally unfair, and should be changed. Some believe that integration should reflect the entire Social Se- curity benefit, not merely the portion funded with employer taxes. Supporters of this change argue that the primary object is to pro- vide a suitable level of retirement income from the two combined sources, regardless of how much of either is funded by employee contributions. This change would increase the allowable amount of integration. Opponents argue that this change would have the effect of reducing benefits for lower-paid workers who need the benefits, and that it is unfair to offset an employee's pension earned under this pension plan by Social Security benefits he had paid for himself. 100 percent offset is needed if the combination of Social Security plus the private pension is to provide the same com- bined replacement ratios at all income levels. But because of the effect of taxes, higher income workers do not need to replace as high a percentage of their earnings as lower income workers in order to maintain their preretirement standard of living. The President's Commission on Pension Policy determined that in 1978 the replacement ratios needed to maintain the preretirement standard of living of single workers was 79 percent for workers earning $6,500, 66 percent for those earning $15,000, and 51 per- cent for those earning $50,000. Probably all agree that the present complex rules need to be sim- plified, but there is little agreement on how to do so. Some want to expand the function of integration rules from merely preventing discrimination to promoting adequate benefits for lower-paid work- ers. A private plan which provides a benefit of 1 percent of the first $12,600 of pay and 38 /2 percent of the excess is, when considered together with Social Security, probably nondiscriminatory. The problem is not discrimination; the problem is that lower-paid work- ers do not receive an adequate pension. Some want to deny tax qualification to integrated plans unless they provide at least mini- mum benefits for lower-paid workers. Some want to eliminate integration entirely, arguing that any in- tegration at all discriminates against the low paid. But elimination of integration would defeat the objective of replacing the income of employees at all wage levels. H. PROTECTING PENSIONS FROM INFLATION 1. INFLATION Inflation has long been part of our economy. The average annual increase in the Consumer Price Index for recent years is as follows:

[In percent] Average Years: Increase' 1940 to 1950 ...... 5.6 1950 to 1960 ...... 2.1 1960 to 1970 ...... 2.7 1970 to 1980 ...... 7.8 19 8 1 ...... 10 .4 19 8 2 ...... 6 .1 1983 ...... 3.2 'Increase before 1981 as reported in Statistical Abstract of the United States, 1982 and 1983: increases for 1981-83 are increased in the annual average CPI-U. This clearly shows that inflation is both continuing and erratic. Any attempts to estimate long-term future inflation are highly speculative. The average American retiring at age 65 has a life expectancy of approximately 17 years. One-fourth of them will live 22 years or longer. The value of $1 after 17 or 22 years under various rates of inflation is shown below:

Valueof $1 Annualrate of inflation(in percent) After17 years After22 years

2 ...... $0.7 1 $0.6 5 3 ...... 61 .52 5 ...... 44 .34 7 ...... 32 .23 10 ...... 20 .12

Even under optimistic assumptions that inflation can be held to 3 percent annually, the average pensioner who receives a fixed pension will see it lose 39 percent of its purchasing power in 17 years and one-fourth will see a 48-percent loss of its purchasing power in 22 years. If higher rates of inflation prevail, fixed pen- sions are devastated. Some have suggested that the increase in the cost of living for the elderly may be less than for the rest of the population. But in the most thorough study of that question ever undertaken, the General Accounting Office found no significant difference between the increases in a Consumer Price Index for retirees and the CPI-U basis now used for the entire urban population.3 The cost of adjusting pensions to keep pace with inflation is ap- proximately as follows:

a U.S. General Accounting Office. A CPI for Retirees Is Not Needed Now But Could Be In the Future. GGD-82-41, June 1, 1982. Washington, D.C. Annual rate of inflation (percent): Cost I 0 ...... $ 10 0 3 ...... 125 6 ...... 156 9 ...... 19 5 I Cost of providing pensions at age 65, adjusted by inflation as a percent of the cost of provid- ing level pensions.

2. INFLATION AND INVESTMENTS High rates of inflation are usually accompanied by high interest rates in both short-term notes and long-term bonds. While upward changes in interest rates cause bond market values to fall, contin- ued high yields over long periods result in high returns on bonds. There has been little if any correlation between common stock prices and inflation, although the total investment return of common stocks over long periods has usually been higher than the return of bonds.

3. HISTORICAL ADJUSTMENT OF PENSIONS FOR INFLATION Social Security, Federal retirement systems, and some municipal retirement systems have adjusted pensions payable by the change in the CPI. The 1983 amendments to the Social Security Act could limit such adjustments to the percentage change in wages of active workers, but this limitation is expected to apply infrequently, if ever. Legislation has been introduced in the Congress which would remove this provision entirely. To my knowledge, no private sector employer has ever promised to provide unlimited CPI adjustments. This partly because of the high cost of such adjustments, but even more because of the unpre- dictability of the costs. No private employer will sign a blank check. A minority of private pension plans, covering perhaps 5 percent of private plan participants, provide limited automatically CPI-in- dexed pensions, with a maximum annual increase usually between 2.5 and 5 percent. In the 1950's and 1960's, a number of equity variable annuity plans were established, under which plan assets were invested in common stocks, and the amount of pension fluctuated with the in- vestment return of the assets, including market value fluctua- tions.4 The hope was that market value increase would pay for pensions that would keep up with inflation. Such plans performed well in the 1950's and 1960's, but declines in market values during the 1970's caused pensions to fall while inflation was soaring, caus- ing many employers to abandon these plans. Approximately two-thirds of large plans gave one or more ad hoc increases to retired members during the last half of the 1970's. Most other large plans and most small plans gave no increases. One study shows that between 1973 and 1979 increases averaged 38 percent of the CPI increase. A separate problem relates to deferred vested pensions. A worker terminating employment at age 35 after 10 years of service may have a right to a deferred pension beginning at age 65 which ap-

4 Johnson, G. E., and D. S. Grubbs. The Variable Annuity. Indianapolis, Research & Review, 1967, rev. 1969. pears to be appropriately related to his compensation and years of service, but inflation of only 3 percent per year will erode 59 per- cent of its value before it even starts. Almost no plan makes any adjustment to deferred vested pensions. 4. CURRENT PoucY ISSUES President Eisenhower stated, "Inflation is not a Robin Hood, taking from the rich to give to the poor. Rather, it deals most cru- elly with those who can least protect themselves." There is wide agreement that inflation seriously erodes private pensions and limits their ability to effectively supplement Social Security as a source of retirement income. Reducing inflation is clearly the larg- est issue. If inflation returns to the 10 percent level, probably no solution is workable. But even if inflation can be held to the 3 per- cent level, a serious problem remains. Some argue that private sector pensions are not able to solve the problem, and therefore an expanded Social Security is the best hope for providing adequate pensions in an inflationary age. While the desirability of providing post-retirement adjustments is recognized, there are no serious proposals to require all plans to adjust all pensions, since employers are not required to provide pensions at all. If substantial additional costs were imposed upon all pension plans, it would discourage the establishment of plans or result in lower benefit levels. For any initial amount of annuity, increasing annuities cost more than level annuities. For any fixed amount of cost, the initial amount of increasing annuity must be less than the initial amount of level annuity that could be provided. No solution can escape these facts. When defined benefit pension plans are terminated, with assets in excess of the value of accrued benefits, the excess may be ap- plied either to improve benefits for participants or to be returned to the employer. Some advocate requiring that part or all of any such excess be applied to increase benefits after retirement. This proposal, with its advantages and disadvantages, is discussed in the larger context of such termination in section K of this chapter. I. PLAN TERMINATION INSURANCE

1. MULTIEMPLOYER PLAN TERMINATION INSURANCE The Multiemployer Pension Plan Amendments Act of 1980 [MPPAA] solved the immediate crisis facing multiemployer plan termination insurance. However it also created great problems. One problem is the large withdrawal liabilities imposed upon withdrawing employers. Employers who entered plans agreeing only to contribute a fixed number of cents for each hour worked by plan participants have been assessed withdrawal liability many times the level of their annual contributions, and in some cases greater than their net worth. A continuing issue is whether such burdens can be lifted, in particular industries or in all industries, without creating greater problems in the process. In many multi- employer plans the withdrawal of individual employers creates no significant problem for the plan. Most multiemployer plans experi- enced withdrawing employers for decades before MPPAA estab- lished withdrawal liability, with no particular problem. But for some multiemployer plans, particularly in declining industries, withdrawal of employers creates a financial crisis and either with- drawal liability or some alternative remedy is greatly needed. MPPAA established two alternative criteria for the actuarial as- sumptions to be used in calculating withdrawal liability, either as- sumptions determined by regulations of the Pension Benefit Guar- anty Corporation [PBGC] or assumptions which are "reasonable" and the actuary's "best estimate" of future experience. Three years later PGBC has not yet issued regulations, forcing plans to use the second criterion. Inevitably, a standard as imprecise as "reasona- ble" has lead to widely divergent viewpoints and a large volume of arbitration. This has resulted in plans and employers spending huge amounts on the arbitration process that could be better spent on benefits and productive uses, if the arbitration could be avoided. If specific standards were established by PBGC or by Congress, this waste could be ended. But there is no agreement on what the standards should be. Opinions about interest assumptions range from current interest rates to the far more conservative interest as- sumptions used for plan funding purposes. A much more serious problem in the long run is the fact that withdrawal liability has created a strong incentive for employers to avoid participating in multiemployer plans. This has already cre- ated strong employer resistance to participating in the plans, and has resulted in many employees not enjoying coverage under multi- employer plans. Many who work closely with multiemployer plans believe that in the long run this resistance poses a very serious threat to the continuation of the multiemployer pension system. Some believe that plan termination insurance for multiemployer plans is needed to protect benefits in plans that terminate. Others recognize this value, but believe that the system is having a grow- ing effect of depriving workers of any coverage at all under the plans and may ultimately destroy the plans. This latter group favors abolishing plan termination insurance for multiemployer plans, or at least substantially modifying it.

2. SINGLE EMPLOYER TERMINATION INSURANCE The author's 1973 analysis of pre-ERISA experience indicated that the unfunded vested benefits under terminated plans averaged about $1 per participant per year, but that if ERISA's provisions had applied about 90 percent of the cost would be paid by contin- gent employer liability, leaving a net cost to PBGC of about $0.10 per participant per year. But the author warned Senator Williams that establishing the plan termination system under ERISA might itself cause a substantial change in this. To be conservative, the premium level was initially established at $1 per participant per year. When this proved inadequate, the premium was increased to $2.60. Now the need is reported in the area of $6 to $8.50. Most who have studied the system expect fur- ther substantial increases in future years unless the system is modified. Part of the problem is that the present system provides incen- tives for some employers to terminate plans with unfunded liabil- ities and to dump the liabilities on PBGC. Proposed legislation en- deavors to end this incentive. Opponents charge that the proposed corrective legislation increases other problems. The system has created difficulties for owners of small businesses who want to sell their business, often upon their own retirement. Buyers are sometimes reluctant to purchase a business because of its contingent employer liability for unfunded pension liabilities. Fear of contingent employer liability, the resulting difficulty of selling businesses, and the burden of growing premiums are disin- centives to establishing defined benefit plans. Since ERISA the es- tablishment of new defined benefit plans has fallen drastically, in part because of plan termination insurance. However, the plan ter- mination insurance has protected benefits under terminated plans, providing pensions to employees that otherwise would have been lost. Observers disagree about whether the benefit or the detriment of plan termination insurance is greater. Some would abolish plan termination insurance altogether, argu- ing that its disadvantages exceed its benefits. Opponents argue that this would deprive employees of needed protection. Some would seek to limit the growth of premiums by limiting the growth of guaranteed benefits. One proposal is to freeze the maxi- mum guaranteed monthly benefit at its present level of $1,602, in- stead of allowing continued annual inflation adjustments. Because most vested benefits are much smaller than this, this change would have only a negligible effect in the early years, but in the long run it might have a significant effect. A second proposal to limit the growth of guaranteed benefits re- lates to plan amendments. No increase in vested benefits resulting from a plan amendment would be guaranteed until 5 years after the amendment is adopted, rather than the 5-year phasein under present law. In addition to reducing the growth in guaranteed ben- efits, this change would substantially reduce the required calcula- tion work on an actual plan termination. Some want to expand the level of guarantees, increasing the $1,602 maximum monthly benefit guaranteed, making the phasein more rapid, or guaranteeing certain nonbasic benefits not now guaranteed. Such changes would increase protection in terminated plans, but also increase the costs of the system. One alternative to limiting guaranteed benefits would require employers to continue funding terminated plans. PBGC would un- dertake the task of guaranteeing benefits only at the point of plan insolvency or certain other events, such as dissolution or bankrupt- cy of the employer. The first of several proposals to strengthen minimum funding re- quirements is to reduce the period over which unfunded liabilities are required to be amoritzed. An alternative approach would change the present law's requirement that the actuary ignore future inflation-indexed increases in the maximum benefit limits; changes which current law has scheduled to occur beginning in 1988. Actuaries are almost unanimous in agreeing that the present law, requiring the actuary to pretend there will be no future infla- tion, is fundamentally unsound and tends to cause underfunding. 3. PLANs TERMINATED DURING THE 1974 WINDow PERIOD All four congressional committees involved with ERISA agreed to some form of the bill by February 21, 1974. It was expected by many in Congress that the committees could resolve their differ- ences and that ERISA could be enacted early in 1974. However, the resolution of differences was protracted and ERISA was not finally enacted until September 2, 1974. But Congress believed that par- ticipants should not be deprived of the protection of plan termina- tion insurance merely because of the delay in passage. Therefore it provided the protection of plan termination insurance to plans which terminated between July 1, 1974 and September 2, 1974, the "window period." It was necessary to determine the date of termination in order to determine which plans properly fall within the window period. It would have been unwise to allow employers to manipulate the date by reciting an "effective date" of termination unrelated to the actual facts of the situation. ERISA provides that, for the purpose of determining the date of termination for this provision, "the corporation shall make the de- termination on the basis of the date on which benefits ceased to accrue or on any other reasonable basis consistent with the pur- poses of this subsection." Congress determined that the date bene- fits cease to accrue would normally be used to determine the date of termination. Before ERISA plans were not required to define the accrued benefit, which ERISA now requires for all plans. In some plans, benefits were not related to years of service and no benefits were provided upon termination of employment before retirement age. In such a plan, it would have been impossible to determine when benefits ceased to accrue. In addition, there could be unusual situations where the date benefits ceased to accrue might somehow distort the underlying facts of the situation, such as a plan providing only past service benefits which is terminated years after the period for which bene- fits are accrued. An alternative basis was needed in such situa- tions. Therefore Congress also allowed use of "any other reasonable basis" for such plans. Some argue it was not the intent of Congress to thereby deny protection of benefits to employees covered under plans under which benefits generally ceased to accrue during the window period, but apparently PBGC disagrees. It is reported that PBGC rejected 200 of the 208 claims for window period coverage. In some of these PBGC disregarded the standard of when benefits ceased to accrue, even though it was clear that participants were continuing to earn continuous service and to accrue benefits after July 1, and instead used the "other reasonable basis" standard to deny protection to participants. The current issue is whether Congress will take corrective action to provide the lost benefits to participants whose benefits ceased to accrue during the window period. A disadvantage of providing the protection is that it will provide cost to the plan termination pro- gram. Proponents of the change argue these are the costs of the protection Congress originally intended to provide. 4. PLANS TERMINATED BEFORE THE WINDOW PERIOD Many pension plans terminated before ERISA under which em- ployees lost promised benefits. Except for the window period plans described above, ERISA provided no relief for participants and beneficiaries who lost benefits under pre-ERISA plan terminations. Some have proposed a "pension losers" bill to ameliorate this loss. Helping the pension losers is obviously a desirable goal, but its costs are a disadvantage. Rather than impose additional cost on the already burdened plan termination insurance program, the bill's proponents would pay for the costs from general revenues. The bill as introduced has serious technical flaws, but most of these could be corrected if agreement were reached on fundamen- tal concepts. A basic problem is lack of plan records. These plans terminated in 1974 and prior, and in many cases the employer is no longer in existence. Records are needed to determine whether a covered plan existed, what classes of employees were covered, what individual employees were covered, the amounts of their accrued benefits, whether they were vested, what benefits they already received from the plan, and other necessary information. In many cases such records are completely lost, or would be difficult and expen- sive to obtain. There are three possible answers to the data problem. All possi- ble solutions involve additional cost. One approach is to limit bene- fits to plans and individuals who can provide adequate evidence to establish their eligibility for a benefit and its amount. This ap- proach would fail to provide benefits to many who actually lost them. A second answer is that taken by some advocates of the proposed legislation. They would apparently make determinations without the benefit of plan documents or plan records. They would assume that all employees of any employer that maintained any defined benefit pension plan were participants in the plan and were vested at the time of plan termination. The bill would assume that all plans provided a benefit of $4 monthly for each year of employ- ment with the employer. It would make other such broad assump- tions. This would result in some individuals receiving benefits who were never entitled to a vested benefit under the terminated plan. Others would receive more or less than their accrued benefits. For most of those who lost benefits, this bill would provide less than the amount lost, since the bill would only replace pensions pay- ments due in future years, not the pensions lost to date during the years since their plans terminated. The third approach is to conclude that there is no practical way to remedy the losses of this group, and that they are part of a much larger group who never had any vested pension benefit to start with. The needs of those without adequate private sector pen- sions could be ameliorated either by increasing Social Security ben- efits or by increasing needs related benefits such as supplemental security income. J. TERMINATION OF PENSION PLAN WITH ASSET REVERSION 1. BACKGROUND There has been an enormous growth of pension plan termina- tions with a reversion of assets to the employer. Some pension plans have assets that exceed the liabilities determined on a plan termination basis. Plan terminations with reversions have in- creased because some employers need the excess assets and realize that, under present law, termination of the pension plan is the only way to get them. Termination of defined benefit pension plans is generally bad for employees. The present law is clear. No withdrawal of plan assets is permit- ted under a continuing plan. Upon termination of a plan, after the liabilities have been satisfied, all of the residual assets may revert to the employer if the plan so provides, except to the extent that any of the residual assets are attributable to employee contribu- tions.

2. REGULATORY ISSUES Thus there are only three regulatory questions under present law. First, what assets are required to satisfy the liabilities under the plan? Purchase of annuities for the accrued benefits clearly sat- isfies the requirement; if the accrued benefit is to be paid off in a lump sum, there is a regulatory question concerning the amount of the lump sum. The second regulatory question, if the plan is contributory, con- cerns the proportion of the residual assets which are attributable to employee contributions. The only remaining regulatory question is whether a termina- tion has actually occurred. The law is clear that, if a termination has actually occurred, 100 percent of the residual assets are recov- erable (except to the extent that any are attributable to employee contributions). If there has not been a termination, none of the assets are recoverable by the employer. Whether or not a termina- tion has occurred is usually clear. But there are two types of purported terminations in which it is not entirely clear whether there has been an actual plan termina- tion. First, under a "spinoff termination" the plan is split into two plans, one for active employees and one for retirees, with all of the excess assets placed in the retirees' plan; the retirees' plan is ter- minated with a reversion of assets to the employer, while the active employees' plan is continued. Second, under a "reestablish- ment termination" the plan is terminated with a reversion of the excess assets to the employer, and immediately another defined benefit plan is established, identical or similar to the plan just ter- minated. In these cases, has there really been a plan termination? A 1963 IRS regulation states: Whether a plan is terminated is generally a question to be determined with regard to all the facts and circum- stances in a particular case. For example, a plan is termi- nated when, in connection with the winding up of the em- ployer's trade or business, the employer begins to dis- charge his employees. However, a plan is not terminated, for example, merely because a employer consolidates or re- places that plan with a comparable plan. Some within IRS have argued that the step transaction doctrine should be applied to spinoff terminations and reestablishment ter- minations. According to the step transaction doctrine, if actions taken in two transactions have substantially the same result as could have been obtained in a single transaction which would have been disallowed, then the two transactions shall be treated as single transaction for tax purposes. They argue that the spinoff ter- mination and the reestablishment termination are merely shams in which the employer dances the two step to accomplish what is es- sentially a withdrawal of assets from an ongoing plan-a reversion that would be clearly prohibited if done in one step. But those inside and outside the IRS have recognized that a seri- ous problem is presented if the step transaction doctrine is applied. If an employer needs the excess assets and can obtain those assets by a complete plan termination, but is blocked from getting them under a two-step termination, the employer will have a very strong incentive to completely terminate the plan. Defined benefit plans are generally good for employees, and their continuation is to be encouraged. Providing employers the incentive to completely termi- nate defined benefit plans would not be good for either employees or the PBGC. This indicates that the step-transaction doctrine should not be applied, regardless of its theoretical justification. Therefore the administration announced that two-step termina- tions would be treated as true terminations, and asset reversions would be allowed if certain conditions are satisfied. Some of the conditions apparently lack any statutory basis.

3. LEGISLATIVE ALTERNATIVES Many feel that present law is inadequate to deal with these prob- lems. A wide variety of legislative alternatives have been suggest- ed. Some have argued against any reversion of assets whatsoever on any plan termination, arguing that all plan assets should in all cases be used to benefit participants and beneficiaries. Opponents argue that such a policy would be unfair to employers, forcing them to forego the benefits of favorable experience even though they are required to make up any losses on unfavorable experience. In addition, a no-reversion policy would discourage employers from establishing defined benefit plans and would encourage employers to fund plans on a marginal basis in order to preclude the exist- ence of excess assets to the extent possible; clearly very negative effects for participants and for PBGC. Others have advocated a modification of this proposal, prohibit- ing all reversions except those which occur under certain circum- stances, such as the bankruptcy of the employer or the termination of employment of all covered employees. This position is based upon the belief that it is generally contrary to the interest of em- ployees to terminate a defined benefit plan and strong disincen- tives should be provided for discontinuing the plan unless the con-

37-569 0 - 84 - 11 tinuation is clearly impossible. While the termination of defined benefit plans is generally against the interest of participants and beneficiaries, this proposal would have the same adverse conse- quences as the prior proposal in discouraging employers from es- tablishing plans and from soundly funding them. A third approach would limit the reversion to the amount of excess rising from particular sources, such as the amount resulting from excess interest income. Such a proposal has technical prob- lems which actuaries can describe in detail. In addition, like any proposals to very substantially limit the reversion of excess assets, it may discourage employers from establishing and soundly funding plans. Another proposal would allocate the excess assets upon termina- tion of a noncontributory defined benefit plan between the employ- ees and employer. Part of the excess assets could be allocated to assure post-retirement increases related to the accrued benefits in order to ameliorate the problems of inflation. The rationale behind this proposal is several fold. First, it is clear that accrued benefits are deteriorated by the effects of inflation and that employees need help with this problem. Second, this approach is a compromise be- tween those who think that all of the excess assets should go to participants and those who think all should go to the employer. Third, the existence of the excess assets results in large part from high interest rates, which in turn are associated with the high in- flation that deteriorates pensions. Thus it is rational that the excess interest earned be used to offset the deterioration in pen- sions from the related inflation. However, this proposal would tend to discourage employers from establishing plans and soundly fund- ing them. Some have considered imposing an excise tax upon any rever- sions which occur. The purpose of this is twofold. First, it would generally discourage terminations of plans for the purpose of ob- taining reversions of assets. Second, it addresses the alleged tax in- equity. It has been pointed out that it is possible for an employer to put money into a pension plan in a year when it is in a high tax bracket and to withdraw it in a later year when it is in a lower tax bracket or when it has no tax liability at all. But the amount re- ceived is included in taxable income. This usually results in cur- rent taxation. If the employer has a tax loss carry over, the distri- bution reduces the amount carried forward and thus generally in- creases future taxes. This is the same tax treatment which anyone receives if he purchases real property for business purposes, claims a tax deduction for depreciation for several years, and then dis- poses of the property at a gain, except that this latter event gener- ally enjoys capital gains treatment as well. It is true that corporations and individuals often time their ac- tions between tax years in such a way as to minimize their taxes. For example, an individual or business may time a tax deduction to occur in a year when the taxpayer is in a higher tax bracket. Under any tax law, law abiding citizens will quite properly conduct their affairs in a lawful manner that tends to minimize taxes. No tax law is perfectly equitable, and it can be argued that such tax planning increases the inequities. But few expect to abolish all in- equities. The tax inequity related to pension plan reversions is a very minor portion of the questions of tax equity related to the timing of events. To my knowledge, tax considerations have not been the primary reason for any plan termination. Like the propos- als to substantially restrict reversions, the excise tax proposal would have the effect of discouraging defined benefit pension plans and of discouraging sound funding. Congress could choose to do nothing and to continue the present law. In favor of this proposal is the fact that it avoids all of the disadvantages sighted for the alternative proposals discurred. It allows employers to recover the excess assets which they need, while encouraging them to continue defined benefit plans by danc- ing the two step. This encouragement of the continuation of plans is a strong argument in favor of present law. But why should the law force employers to dance the two step? If the result is right, should it not be allowed in one step, permitting such withdrawal from ongoing plans without the farce of pretend- ing there has been a termination? The second problem of current law is that the withdrawal of "excess" assets, which are "excess" only when viewed from the perspective of plan termination but are really needed for funding an ongoing plan, does undermine the minimum funding requirements and tends to undermine sound funding. Third, under present administrative interpretations the law forces plan sponsors to purchase annuities under ongoing plans an approach which may increase the cost of funding the plan in the long run. Fourth, this approach abolishes any cushion of pro- tection for PBGC. Finally, the approach taken by the regulators under present law may not withstand challenge in the courts. Another proposed solution is to allow loans of excess assets from ongoing plans, to be paid back with amortization over some stipu- lated period. Like present law, this has the advantage of allowing employers to get the assets which they need for current purposes and encourages them to continue their defined benefit plans. Unlike present law, the repayment provision would gradually re- store the sound funding position of the plan. But this approach in- volves greater tax inequities than present law and would be subject to potential tax abuse. Under current law the reversion is taxable income to the employer, while under the loan approach the amount withdrawn would not be taxable income. Under the loan approach an employer might put money into the fund on a very strong fund- ing basis, contributing the maximum deductible limit on the basis of very conservative methods and assumptions in order to maxi- mize the tax deduction; it could then withdraw the money by way of a loan with no current taxable income, even if the corporation is in the highest tax bracket currently. ERISA and the Internal Revenue Code could be amended to allow employers to make withdrawals of the excess assets from on- going plans, thus removing the incentive to terminate the plan. This withdrawal would put the plan in a funding position that is consistent with the criteria under present law of the alternative minimum funding standard. In order to protect the security of ac- crued benefits, the amount to be withdrawn could be limited in three ways. First, since a withdrawal of assets weakens the funding of the plan on an ongoing basis, the amount withdrawn could be treated as an actuarial loss, the same as if the assets had declined in value by this amount. All actuarial losses are required to be am- ortized over 15 years. Thus the amount withdrawn to meet immedi- ate needs would be rapid by the employer over a reasonable period. Second, the assets withdrawn would be limited to assets in excess of the value of all accrued benefits, so that all accrued benefits will remain funded. Third, a cushion of assets in excess of the value of accrued benefits could be required to be left behind, to make im- probable the chance that adverse fluctuations shortly after the withdrawal will cause the accrued benefits to be less than fully funded. The required cushion should not be unnecessarily large, since that would give employers a greater incentive to terminate the plan to recover the entire excess assets. The fundamental problem is that current law gives employers an incentive to terminate defined benefit plans in order to obtain the plan's excess assets. Administrative interpretations of current law allow employers to bypass the law's nonreversion provision, to the detriment of sound funding. A solution is needed to allow employ- ers to withdraw the excess assets without terminating the plan, while maintaining controls to protect the funding of benefits. K. INVESTMENT OF PENSION PLAN ASSETS

1. BACKGROUND Investment results of plans have enormous effect upon plan ben- efits and upon employer costs. Under all multiemployer plans and all defined contribution plans the investment return increases or decreases the plan benefits that can be provided. Under single em- ployer defined benefit plans the investment return directly affects the employer's costs, but indirectly affects the employer's decision to establish or discontinue the plan or to increase or decrease bene- fits. Investment return also affects the costs of plan termination in- surance. The most important investment issues are those faced by individ- ual plan sponsors, investment managers, and in some cases plan participants, rather than the Federal Government. For example, no one has suggested that the Federal Government should determine the proportion of plan assets to be invested in common stocks, a decision apt to have far more effect on plan benefits and costs than any decision made by Congress. ERISA did codify fiduciary standards, however. It also prohibited certain transactions, and established a procedure for granting ex- emptions from the prohibited transactions.

2. PROHIBITED TRANSACTIONS ERISA generally provides that all transactions between interest- ed parties are prohibited unless specifically exempted. Every pro- hibited transaction is penalized by an automatic 5 percent excise tax. The prohibited transaction requirement may actually deter some investments that are in the plan's best interest. While the exemp- tion process is available to allow party-in-interest transactions which the Department of Labor determines to be harmless, the time required to obtain an exemption may preclude use of the proc- ess, and the expense involved either increases the cost of operating the plan or deters use of the exemption process. Some advocate replacing the prohibited transaction standard with the arm's length standard formerly applicable to pension plans. Under this approach the plan can enter transactions with a party-in-interest if the transactions are comparable to those a fidu- ciary would conduct with a stranger. For example, the plan could lend money to the employer if the security, interest and other terms are as favorable to the plan as a comparable loan a bank would make to the employer, and as favorable to the plan as other investment opportunities. If the arm's length standard is followed, it will well serve the interests of participants. Opponents to a return to the arm's length standard say that its problem is the difficulty of supervision and enforcement. They point to actual abuses that occurred in party-in-interest transac- tions before ERISA. The issue involves trying to weigh the additional burden of the prohibited transaction requirement against the volume of abuse under the arm's length standard. A middle course would be an arm's length standard with reporting of all party-in-interest trans- actions to the Department of Labor for after-the-fact review. Another problem area is the automatic nature of the 5 percent excise tax on prohibited transactions. Most prohibited transactions are inadvertent and are harmless. It has been suggested that the IRS should have the authority to waive the excise tax if the IRS finds the transaction inadvertent and harmless. There are ques- tions of whether the magnitude of the problem justifies the addi- tional governmental administrative expense, and whether the availability of the. waiver might make plan administrators and others careless. 3. INCENTIVE COMPENSATION OF INVESTMENT MANAGERS It is not clear whether ERISA prohibits compensating invest- ment managers on the basis of investment performance. This may depend on whether the compensation is paid from plan assets or paid separately by the employer. Some believe it would be in the interest of participants to com- pensate investment management on the basis of investment per- formance. Providing more compensation for superior performance may cause the investment manager to work harder and produce better results. On the other hand, providing incentive compensation may result in overcompensation of the investment manager, thus wasting plan assets if the compensation is paid from the plan. This is particular- ly true to whatever extent favorable return is the result of chance variation. Incentive compensation may provide the wrong incentive, en- couraging the manager to take excessive risks in order to increase his own compensation. Whether incentive compensation should be allowed, or the condi- tions under which it would be allowed, will be a continuing issue. Some would prefer any clarification to the existing uncertainty. 4. "SOCIAL INVESTMENT" Some forces in organized labor, as well as some endeavoring to support broad public interests, argue that the investment of pen- sion plan assets should serve social purposes as well as the usual investment goals. Organized labor often sees a connection between jobs for union members and the investment of pension plan assets. For example, it is argued that the union-organized textile industry of the North- east is losing out to new, lower-cost, nonunion textile factories in the South, and that the new factories are financed in part by the investments of pension assets of plans for Northeastern union workers. Unions would prefer that the assets instead be invested in the development of business that produce jobs for their members. Some multiemployer plans have invested in local construction projects using union labor. Another example is the investment of the pension assets for New York City employees in loans to the city in 1975. The city needed the loans to prevent a bankruptcy that, among other things, would have cost many of the covered workers their jobs. It was argued that the loans were imprudent investments for the pension funds, since the Nation's leading banks had refused to extend comparable credit because they perceived the risk as too great. Some want particular plans to make investments that will direct- ly benefit participants in those plans, such as new housing or a children's day care center in the area where the particular workers live. Others want pension funds invested in areas they believe serve broad social purposes, such as low-cost housing. Another aspect of so-called "social investing" is the blacklisting of invest- ments that are considered socially undersirable. These may include companies doing business in South Africa, or companies with an antiunion reputation. There is by no means agreement on what in- vestments are or are not socially desirable. Young workers may be most interested in expanding jobs, while retirees may be primarily concerned with safe investments to protect their pensions. Most employers and banks have argued that plan assets should be invested solely in the interests of plan participants and benefici- aries, and that social investment objectives may either increase in- vestment risk or decrease investment return. Blacklisting certain investments has little effect on the investor or on the business being blacklisted. From the viewpoint of the pension fund, many good alternatives for investment are available, so eliminating one alternative does not matter. Opinions differ on whether there is an adverse effect from eliminating a broad class of investments, such as all companies doing business in South Africa. If the business being blacklisted is actually a good investment, it will have no trouble finding other investors at comparable terms. Positive in- vestments for social purposes, however, may adversely affect a plan if it causes the plan to accept higher risks or lower investment re- turns than are otherwise available. But other socially designed in- vestments may have risk levels and returns as good as other avail- able investments. A related question is who should control investment policy. Tra- ditionally, investment policy has been controlled by plan sponsors, either employers or the trustees of multiemployer plans, but this control was often delegated to the investment manager. Increasing- ly unions are demanding a voice in investment policy under single employer plans. The key policy question from the government's viewpoint is whether, or under what conditions, investing for social purposes should be allowed or prohibited as a violation of fiduciary stand- ards. L. FEDERAL ADMINISTRATION OF PENSION POLICY

1. BACKGROUND When ERISA was being developed, the issue was raised as to whether it would be better to have pension plans regulated by the Department of Labor, the Internal Revenue Service, or some new third body designed to regulate pension plans. In large part as the result of interests in protecting turf, ERISA provided for all three. As a practical matter it frequently appears that there are four, since the Department of the Treasury and the Internal Revenue Service have often seemed to operate separately. In addition, the Securities and Exchange Commission, the Department of Defense, the Department of Justice and other agencies have regulated par- ticular aspects of pension plans.

2. COORDINATION PROBLEMS One problem has been the lack of coordination. This has resulted in extended delays, a waste of Government manpower, conflicting policies and ineffectual management. Fortunately the extreme problems in this area that followed enactment of ERISA have been resolved, but problems continue. The most recent major example was the extended delay in devel- oping the administration's policy regarding the termination of overfunded plans with a revision of assets to the employer. Months of delay and conflicting public statements left plan sponsors who were anxious to follow the law unable to act at all. One major cor- poration, frustrated by doubts that the problem would ever be re- solved satisfactorily, terminated its pension plan covering thou- sands of employees, although the final decision of the administra- tion would have made this unnecessary. When the agencies failed to agree, the matter was referred to the Cabinet Council on Eco- nomic Affairs for resolution. Another classic case of lack of coordination was the investigation of the Central States Teamsters plans by the Department of Labor, the Internal Revenue Service, and the Department of Justice. The Senate Investigations Subcommittee found that an amazing ab- sence of coordination between the agencies adversely affected en- forcement. The problems relate not only to coordination between the agen- cies but to coordination within each. In the Department of Labor, responsibilities have been divided between the Office of Pension and Welfare Benefit Programs [PWBP], the Office of the Solicitor, and the field offices of the Labor Management Standards Adminis- tration. When ERISA was enacted, Congress specifically recognized the IRS administration of pension plans had been poorly coordinated, and for the first time specifically legislated the creation of an As- sistant Commissioner to coordinate IRS policy. Unfortunately, the new Assistant Commissioner was given no direct authority over the field operation, which was under the district directors. Regulations were developed by the office of Chief Counsel, who reports to the General Counsel of the Department of the Treasury. All regula- tions, revenue rulings, revenue procedures and many other items were subject to approval of the separate Office of the Tax Legisla- tive Counsel of the Department of the Treasury. In both IRS and the Department of Labor, lack of internal co- ordination has been a major problem. Within the PBGC, organized for a single purpose, internal coordination has not been a signifi- cant obstacle. These comments are not intended as a criticism of any of the dedicated and capable individuals who have conscientiously strug- gled to fulfill their responsibilities within this bureaucratic jungle.

3. PROBLEMS OF ROLE Lingering uncertainties remain concerning the roles assigned to the various agencies. The PBGC often appears to have perceived its primary role to be an insurance company, rather than a protector of pensions. Within the Department of Labor pension plans have usually been seen as one of a number of concerns regarding labor- management relations, although fortunately in recent months the Director of PWBP has significantly improved that situation. The IRS has viewed pensions primarily from its role as tax collector. Each of these agencies is carrying out its responsibilities to pen- sion plans within the context of its primary responsibility. None has been assigned the responsibility of encouraging the establish- ment and maintenance of pension plans. Nor is there anywhere in the administration any one who has been assigned retirement income policy as his primary responsibility. In addition, there is no one with the wider role of responsibility for retirement income policy encompassing both Social Security and private pension plans.

4. PROPOSALS To SOLVE THE PROBLEM Legislation has been introduced to establish a single agency to regulate pension and welfare plans, combining tasks now divided between the Department of Labor, IRS, and PWBP. An alternative proposal would enable a coordinating body to establish policy and coordinate efforts. It has been suggested that congressional coordination of its con- cerns in this area might be enhanced by establishing a Joint Com- mittee on Pension and Welfare Plans, operating somewhat similar- ly to the Joint Committee on Taxation, to provide strong staff sup- port and coordinate activities of the four congressional committees concerned with this area. M. FEDERAL EMPLOYEE RETIREMENT PLANS 1. BACKGROUND

(A) COVERAGE Retirement systems cover over 5 million Federal employees and 4 million retirees and beneficiaries. The two largest system are the civil service retirement system [CSRS] and the military retirement system, but others are also significant.

(B) CIVIL SERVICE RETIREMENT SYSTEM CSRS covers 2.7 million Federal employees and approximately 2 million retired employees and beneficiaries receiving payments. Federal employees hired before 1984 were not covered under Social Security for their Federal employment. However about two-thirds of CSRS retirees actually received Social Security benefits for other employment. Full benefits are payable if any one of these age and service re- quirements are met: (1) Age 55 and 30 years of service. (2) Age 60 and 20 years of service; or (3) Age 62 and 5 years of service. Employees who terminate their service after 5 years of service are vested in deferred pension beginning at age 62, if they do not withdraw their contributions. Pensions equal a percentage of the employee's average pay for the three highest consecutive years. The percentage is 1.5 percent for each of the first 5 years, 1.75 percent for each of the next 5 years, and 2 percent for each year over 10. This results in a pen- sion of 56.25 percent after 30 years or 76.25 percent after 40 years. This is a maximum of 80 percent after 42 or more years. After pen- sions begin, they are adjusted annually by the same percentage as under Social Security, generally equal to the cost-of-living increase for the prior year. CSRS also provides benefits upon disability or to survivors upon death. Employees hired before 1984 contribute 7 percent of pay to CSRS. The Federal Government pays the balance of the cost. The Government now contributes 32 percent of pay, which will eventu- ally grow to 37 percent, a percentage expected to be sufficient to fund the system in the long run. But accounting under the unified budget includes a much smaller amount, thus understating the cost of the system. The 1983 amendments to the Social Security Act require Federal employees hired after 1983 to be covered under Social Security. It was generally understood that CSRS would be substantially modi- fied for the post-1983 employees, reflecting their contributions and benefits under Social Security. Congress is currently studying this problem. As an interim measure Congress decided that new Feder- al employees would contribute 7 percent of pay to Social Security plus 1.3 percent to CSRS, since the total 8.3 percent would be equal to the total amount contributed to the combined systems by Feder- al employees hired before 1984. (C) OTHER FEDERAL CIVILIAN RETIREMENT SYSTEMS Other retirement systems cover special groups of Federal civilian employees. Some of these are more generous than the CSRS. For example, the foreign service retirement and disability system has earlier retired ages and provides larger pensions than CSRS.

(D) MILITARY RETIREMENT SYSTEM Military retirement system applies to military employees of the Army, Navy, Marine Corps, and Air Force. Separate but similar plans cover uniformed employees of the Coast Guard, the Public Health Service, and the National Oceanic and Atmospheric Admin- istration. It provides benefits upon retirement from active duty after 20 or more years of service, regardless of age. The amount of pension is 2.5 percent of pay times years of service, with a maximum of 75 percent after 30 or more years. The pension is calculated based on the final rate of basic pay for those who became members before September 8, 1980, and on the 3-year-final average basic pay for those hired later. Basic pay represents approximately 76 percent of total pay for a typical retiree. After benefits begin, they are indexed annually by CPI changes. Disability benefits and partially subsidized survivor benefits are also provided. The valuation for fiscal year 1983 shows that the annual cost of the military retirement system, including the normal cost and 75- year funding of unfunded liabilities, is 79 percent of pay. As with the CSRS, the unified budget has the effect of understating the true cost of the system.

2. NEW CIVILIAN EMPLOYEES UNDER CSRS The immediate issue that must be resolved is what contributions and benefits should apply to Federal civilian employees hired after 1983. There are many considerations in the determination, and ac- cordingly many proposed solutions.

(A) COMPARABILITY WITH PRE-1984 EMPLOYEES Some argue that the aggregate contributions and benefits under Social Security and CSRS for employees hired after 1983 should be the same as for those hired before 1984. This approach has already been used in the interim measure now providing total employee contributions of 8.3 percent. Under this approach, CSRS benefits should be the same as at present, reduced by whatever benefits are payable from Social Security attributable to Federal employment. For an individual with both governmental and nongovernmental coverage under Social Security, there are a variety of theories on how to determine the portion of Social Security benefits attributa- ble to Federal Government employment; the different approaches produce very different results. Some who subscribe to the general rationale of providing benefits comparable to pre-1984 employees would develop approximate formulas to accomplish this under CSRS, without a direct offset for Social Security benefits. (B) COMPARABILITY WITH PRIVATE SECTOR PLANS Some argue that new Federal employees should receive benefits comparable to those typically provided to employees in the private sector. The private sector has great variety, but in general retire- ment programs, together with Social Security, are much less liberal than under CSRS, particularly regarding retirement age and post- retirement inflation adjustments. Also, the cost of private sector plans plus Social Security is much lower than for CSRS. Some argue that this fails to reflect the fact that private employers commonly provide profit-sharing plans or other fringe benefits not provided to Federal employees. This is true, but even when these other benefit costs are added, private sector costs are much lower. This rationale leads to a pension benefit design comparable to that of a major corporation, perhaps with a supplemental defined contribution plan (discussed separately below). Some argue that comparability must weigh the entire compensa- tion package, including salaries, retirement benefits, death bene- fits, health care benefits, vacations, holidays, and other fringe ben- efits. This rationale justifies providing Federal employees higher, equal, or lower pensions than their private sector counterparts, de- pending on whether the nonpension compensation of Federal em- ployees is lower, equal or higher than private sector employees. There is no agreement on the relative levels of nonpension compen- sation.

(C) EMPLOYEE NEEDS A third approach is to establish a program primarily designed to meet employee needs, regardless of whether that is comparable to either the present CSRS or the private sector. This approach raises fundamental questions concerning when employees need to receive benefits and how much benefits they need. At retirement age, the employee who has spent his entire career with an employer ideally should receive an amount of retirement income from all sources combined sufficient to allow him to contin- ue his preretirement standard of living. The President's Commis- sion on Pension Policy studied the question of how much of gross compensation needs to be replaced to accomplish this, taking ac- count of taxes, reduction in work related expenses, and savings. The Commission reported that replacement ratios ranging from 86 to 51 percent are needed, depending on the retiree's marital status and level of preretirement gross income. Part of the needed income will be provided by Social Security. The balance could be an ideal target for the CSRS. Some would argue that part of this balance should be provided by individual savings, but others would assert that, as a practical matter, few lower income employees are able to accumulate any substantial savings. The questions of when benefits need to begin and the extent to which inflation adjustments are needed are discussed below. (D) RETIREMENT AGE CSRS now allows long service employees to retire on unreduced pensions at age 55. Most private sector plans provide unreduced pensions at 65, with actuarially reduced pensions for those retiring earlier to reflect the greater cost of starting pensions earlier. Some private plans provide unreduced benefits before 65. Social Security has historically provided full pensions at age 65 and reduced pensions as early as age 62. 1983 amendments to the Social Security Act will increase the normal retirement age from 65 to 67, phasing in the increase over a period of years. Some be- lieve that, since new Federal employees will receive part of their benefit under Social Security and part under CSRS, the normal re- tirement age under CSRS should be the same as under Social Secu- rity. But those who believe new Federal employees should be treat- ed like older ones argue that the CSRS retirement age should con- tinue to be age 55 after 30 years of service for new employees, and that CSRS should pay supplemental benefits between age 55 and the age Social Security benefits start, comparable to the "30 and out" benefits in the auto industry. At what age are individuals no longer able to perform their jobs satisfactorily? This varies greatly from individual to individual. Past experience under Social Security shows that most American workers have actually worked to age 65 until recent years. Since health is improving, some believe that the capacity to work longer will also increase. Congress determined that 67 is a generally ac- ceptable retirement age under Social Security. In recent years, however, most workers retiring under Social Security have retired between age 62 and 65. Although the decision to retire earlier is sometimes dictated by declining health or unavailability of work, clearly many Americans desire to retire early. A retirement plan with earlier retirement ages may be more ex- pensive than one with later retirement ages.

(E) ADJUSTMENTS FOR INFLATION Inflation erodes the purchasing power of pension income. If pen- sions are not adjusted for inflation, the retired person may be gradually reduced to poverty. A recent study by the Bureau of Labor Statistics indicated that the percentage change in the cost of goods and services used by retirees was almost identical to that for active workers. This leads to the conclusion that CSRS benefits should be fully adjusted for changes in the CPI (subject to the same limitations that apply to Social Security), just as at present. Full indexing for CPI increases is very expensive. Although the majority of private sector retirees receive some type of adjustment for inflation, no private sector plans match 100 percent of CPI in- creases. Those who advocate comparability with the private sector argue that Federal employees should not be treated better than private sector employees. In addition, some argue that full indexing cannot be afforded. (F) DEFINED CONTRIBUTION PLANS All pension plans are either defined benefit plans or defined con- tribution plans. Which type should be used for Federal employees, or is a combination best? CSRS, like most pension plans, is a defined benefit plan. This means that the plan's benefit formula defines the amount of pen- sion to be paid, e.g., 56.25 percent of pay after 30 years of service. The amount of pension does not depend on the plan's investment experience. The employer contributes whatever amounts are needed, in addition to employee contributions, to provide the de- sired benefits. In contrast, a defined contribution plan (sometimes called an "in- dividual account plan") establishes an individual account for each employee. The plan defines the amount of employee contributions and employer contributions to be added to each employee's ac- count. The contributions are usually invested in a trust fund. Each employee's account receives an allocation of the trust's investment income, gains and losses. Many defined contribution plans allow the individual employee to direct part or all of his account into in- vestment funds with differing investment objectives, such as a money market fund and a common stock fund. The benefits paid from a defined contribution plan are whatever can be provided by the account balance. Many plans pay the ac- count balance in a lump sum. Others apply the account balance to provide an annuity payable for the lifetime of the retired employ- ee, or for the joint lifetimes of the employee and the employee's spouse. Some plans allow the employee a choice between receiving his account balance as a lump sum or as an annuity. Under many defined contribution plans employee contributions are voluntary. Under some of these, called thrift plans or savings plans, employer contributions are determined as a percentage of employee contributions. Under some types of defined contribution plans (defined in Inter- nal Revenue Code sections 401(k), 403(b), and 457), employee contri- butions are excluded from the employee's taxable income, and thus reduce his current income tax. Defined contribution plans are very popular with employees. If the investment income is favorable, they provide substantial sav- ings, which can be used to meet retirement needs. But as a method of providing retirement income, defined contri- bution plans have serious shortcomings. Because the benefits depend on investment return, the amount of benefits is unpredict- able. If the plan allows the employee to receive his account balance in the form of a lump sum, he may spend it or invest it poorly, and be left in poverty in his old age. Many employers believe that a defined benefit plan is needed to assure a basic level of retirement income, but that a defined contri- bution plan can provide a desirable supplement. The extent, if any, to which employer resources should be divert- ed from providing defined benefits in order to fund a defined con- tribution plan is arguable. Even if no employer contributions are used, a defined contribution plan could be established to receive voluntary employee contributions. As long as such employee contri- butions are already excludable from taxable income for private sector employees and for State and local government employees, there appears no reason to bar similar treatment for Federal em- ployees.

(G) EMPLOYEE CONTRIBUTIONS What, if any, employee contributions should be paid under CSRS by employees hired after 1983? Employee FICA taxes for Social Security and medicare total 7 percent of pay in 1984, and will increase to 7.65 percent by 1990. A special 0.3 percent credit applies in 1984. The CSRS requires 7 percent employee contributions for employ- ees hired before 1984, and these employees must also pay 1.3 per- cent for medicare, a total of 8.3 percent. As a stop-gap measure, Congress required Federal employees hired after 1983 to contribute 1.3 percent of pay to CSRS. Now Congress must decide what, if any, rate should apply in the long run. The principal advantage of employee contributions is that they reduce the level of employer contributions needed to fund the plan. Disadvantages include both the cost to employees and an increase in administrative work. Most private employer pension plans do not require any employee contributions. However many employers maintain sup- plemental defined contributions plans to which employees may con- tribute.

(H) INTEGRATION WITH SOCIAL SECURITY Integration with Social Security benefits related to several of the above issues. Some of the possible alternatives are: (i) Offset by 100 percent of the Social Security benefit; (ii) offset by part, perhaps 50 percent, of the Social Security benefit; (iii) no offset; and (iv) a step- rate benefit formula (higher benefit percentage above a certain pay level) to approximately integrate with Social Security. The amount of any offset and the basic benefit formula will be considered together. A lower basic benefit formula may be adopted if there is no offset than if there is an offset. A 100 percent offset might bring complaints from the private sector, which is not permitted to have a 100 percent offset.

3. CSRS BENEFITS FOR EMPLOYEES HIRED BEFORE 1984 The 1983 amendments to the Social Security Act clearly require changes to CSRS for employees hired after 1983, but there is no comparable requirement to change CSRS for those hired earlier. For those hired before 1984, CSRS can continue unchanged. Nevertheless some urge reevaluation of CSRS for pre-1984 em- ployees. If different benefits are provided for those hired later, would not the rationale leading to the new program, as well as comparability for the two groups, suggest changes for pre-1984 em- ployees also? Many expect the program for new employees to be less liberal than that for older employees. Some argue that older employees have legitimate expectations regarding their benefits, and in many cases have continued their Federal employment because of them. Some of them have already planned their retirement. This argu- ment is particularly strong regarding long service employees close to retirement. Some argue employees have a legal contractual right that Congress cannot change, but others reject this legal argument (except with respect to judges). Others argue that legitimate expectations apply only to benefits already accrued for service to date. Employees may hope the pro- gram will be continued, just as they hope that their jobs will con- tinue. But no Federal employee is guaranteed that his job will not be reclassified to a lower salary grade, or abolished entirely. This rationale leads to grandfather clauses to protect benefits accrued to date, but does not preclude the possibility of reduced benefits ac- crued for the future. Those arguing for change are primarily con- cerned with reducing the cost of CSRS. If changes are to be made for employees hired before 1984, almost all of the considerations applicable to new employees apply to those hired earlier.

4. OTHER FEDERAL SYSTEMS FOR FEDERAL EMPLOYERS Almost all of the issues related to CSRS apply to other programs for Federal civilian employees. In addition there are supplemental issues. Fundamental is the question of whether each of the other pro- grams should in fact differ from CSRS. Do the differences reflect significant differing needs of the employer or the employee, or do they reflect inequities stemming from historical roots which are no longer relevant? Are there compelling reasons not to combine the programs? What are the cost considerations? One significant need for difference is in retirement ages for occu- pations requiring superior physical condition, such as security posi- tions. For these, consideration could be given to transferring the in- dividuals to physically less demanding positions rather than retir- ing them, but if this is not feasible differing retirement ages may be needed.

5. MILITARY RETIREMENT SYSTEM Some have suggested that the military retirement system should be modified. The benefits are substantially more liberal than even CSRS, and as a result it is much more expensive. Most of the con- siderations applicable to CSRS apply to the military system, but there are other considerations as well.

(A) PERSONNEL PURPOSES OF THE PRESENT SYSTEM The present system is an important management tool for the uniformed services. It is not helpful in recruiting new members, but it is a major retention tool for the individuals with 5 or 10 years of service, as clearly shown in the Fifth Quadrennial Review of Military Compensation [QRMC], released in January 1984. Also, the retirement age provisions may service various other personnel purposes. (B) RETIREMENT AGE The military retirement system provides unreduced pensions upon retirement from active duty after 20 years of service, regard- less of age. Thus an individual entering at age 18 can begin receiv- ing a pension at age 38. Consideration must be given to the personnel needs of the mili- tary services. Many jobs require individuals in vigorous physical condition. Other jobs do not require unusual physical condition. The Government might consider how many of its military jobs re- quire unusual physical condition, and at what age individuals gen- erally become unsuited for them. It might consider whether there are other Government jobs the individuals could perform when they are no longer able to carry mortar shells. Such a study might lead to establishing a normal retirement age the same as CSRS, with unreduced early retirement benefits avail- able only for individuals with 20 or more years of service for whom the Government has no suitable job. Availability of retirement at early ages may cause the services to lose some valuable members at the peaks of their careers. But the early retirement ages may help create openings for promotion, and may be related to an "up-or-out" policy. If so, early retirement ages could be specifically limited to those discharged under the up-or-out provision, with substantial cost saving.

(C) IMMEDIATE OR DEFERRED PENSIONS In the private sector, an employee who terminates employment at age 40 after 20 years of service usually is entitled to a deferred vested pension beginning at age 65. This is considered appropriate, since the individual is expected to take another job, accrue addi- tional pension benefits and retire at 65 with the combined pensions earned from all his employment. The military retirement system is based instead on providing the 40-year-old an immediate pension. This difference is a key design issue.

(D) COORDINATION WITH SOCIAL SECURITY Members of the military services are covered under Social Secu- rity, and receive full Social Security credit for their military serv- ice. Thus, an individual who enters the military service at age 20 and retires at age 50 with a pension of 75 percent of basic pay, ad- justed for future cost-of-living increases, can also begin collecting a substantial Social Security benefit at age 62 or 65, based on the same service and earnings. Neither pension is explicitly offset or adjusted for the other. An implicit partial offset results from the fact that only basic pay is considered in Social Security earnings. Thus, the Social Security benefit reflects less than full military pay. Possible alternatives can be considered.

(E) RESERVE MILITARY PERSONNEL Consideration can also be given to military retirement system provisions related to reserve military personnel. (F) TRANSITION Present military personnel have expectations regarding their ac- crued and projected benefits. If any change is made, careful atten- tion must be given to the extent and method of preserving these expectations.

6. FUNDING AND ACCOUNTING FOR FEDERAL PLANS The contributions made to the various Federal retirement sys- tems have been, and in some cases still are, generally far less than the accruing costs. Compared to private sector plans, assets repre- sent a very low proportion of accrued liabilities. Some have proposed that all Federal pension plans be funded on a sound basis, under standards similar to those required for private sector plans. Others argue that funding is not needed since the full faith and credit of the United States stands behind the pension promises and, unlike private plans, the permanence of the employer is assured. In addition, some argue that funding a Federal plan has no sub- stance. An unfunded liability for pensions is an obligation of the Government. To fund it, the Government can issue a bond to borrow the money needed and contribute it to a trust, which in turn would purchase comparable Federal bonds. One Federal obli- gation would merely be substituted for another. Some argue that the Federal budget as a whole, and the budgets of individual agencies, should show pension costs on an accrual basis. Budgets now understate accruing pension costs. This may distort cost comparisons used to determine whether to perform cer- tain jobs internally or contract out for them. However, the rest of the Federal budget is not on an accrual basis. Unlike a corporation, if the Government builds an office building or an ICBM this year, it does not amortize the cost over the expected lifetime. N. PENSION PLANS FOR STATE AND LOCAL GOVERNMENT EMPLOYEES 1. BACKGROUND About 6,600 retirement systems cover over 10 million employees of State and local governments, and provide benefits to over 2 mil- lion retirees and beneficiaries. 5 The plans are so diverse in their size, design, funding and other characteristics that no brief descrip- tion of them is possible. These governmental plans are subject to part of the qualification requirements of the Internal Revenue Code. But many governmen- tal plans have ignored the requirements of the code, and enforce- ment by IRS has been nil. In most cases, however, the noncompli- ance has not harmed participants and beneficiaries. Government plans are specifically exempt from title I of ERISA, including the reporting and disclosure requirements and the fiduci- ary requirements. Governmental plans are also exempt from the

5 U.S. Congress, House of Representatives. Committee on Education and Labor. Pension Task Force Report on Public Employee Retirement Systems. 96th Cong., 1st Sess. Washington, U.S. Govt. Print Off., 1978. participation, vesting, funding, joint and survivor, and various other requirements of both title I and title II, and the plan termi- nation insurance provisions of title IV. State and local government plans face all of these issues of plan design faced by any pension plan, such as retirement age, benefit formula, etc. However these are not generally considered to be concerns of the Federal Government, and therefore will not be considered here. This chapter only considers the continuing issues of whether Feder- al laws similar to those applicable to private sector pension plans should apply to plans of State and local governments.

2. ROLE OF THE FEDERAL GOVERNMENT If legislation is needed regarding governmental pension plans, the question arises whether Federal or State legislation is prefera- ble. There are valid arguments for both. A second question concerns the legal power of the Federal Gov- ernment to regulate State pension plans. In National League of Cities v. Usery (426 U.S. 833), the Supreme Court held that the Fair Labor Standards Act could not impose wage and hour requirements on State and local governments, because it interfered with "the conduct of integral governmental functions," such as scheduling overtime by police and firefighters. But many think that Federal regulation of pension plans which did not impose substantial addi- tional costs would be allowed by the court. Perhaps regulation of reporting, disclosure, and fiduciary requirements would be disal- lowed because of their cost. There are also practical political questions concerning any Feder- al effort to impose standards on State and local governments.

3. REPORTING AND DISCLOSURE REQUIREMENTS Some governmental plans have done an excellent job of reporting and disclosure, while others have done almost nothing. Particularly troublesome are those plans that have failed to provide employees with any description of their benefits comparable to ERISA's sum- mary plan description. Some have proposed subjecting all public employee plans to the same reporting and disclosure requirements as apply to private em- ployer plans. Others believe that parts of those requirements, such as summary plan descriptions and statements of accrued benefits, would help employees while other parts would merely add expense without helping any employee to understand or receive the benefits he is entitled to. Some advocate adopting Federal standards, but exempting States that adopt acceptable State standards.

4. FIDUCIARY REQUIREMENTS Some have proposed subjecting public employee plans to ERISA fiduciary requirements, arguing that there is no reason to exempt public employee plans for the same high standards. This would create practical problems if the Department of Labor had to rule on prohibited transaction exemptions for States, a situ- ation some would like to avoid. Some would prefer to apply the arms length standard to public employee plans rather than the prohibited transaction standard. Others argue that these plans are already subject to the prohibited transaction rules of the Internal Revenue Code. There has been relatively little abuse in this area, and it is not clear that the proposed changes in the law would actually increase the benefit of any participant.

5. FUNDING REQUIREMENTS Many public employee plans are soundly funded, but some are badly underfunded, creating doubts about whether some of the plans will be able to pay all of their promised benefits. Some believe that minimum funding requirements should apply to public employee plans. Some doubt that such requirements would be constitutional. Because such requirements would force large increases in current pension contributions by some Govern- ment employers, strong opposition would be expected. Some think that establishing reporting requirements for public employee plans would have an indirect effect of encouraging stronger funding. But because the proposed reporting requirements would in some cases show accrued liabilities substantially lower than those determined under strong actuarial cost methods, they might actually cause a weakening of funding. 0. CONCLUSION This chapter has endeavored to address the more immediate issues facing pension plans, but by no means all of the issues. Issues related to Social Security itself were not addressed, although these are very closely related to private pension issues. The two programs together must be correlated to meet the needs of retired workers. Any change in one may create a need for changing the other. Every issue that affects employment, inflation, investment re- turns or life expectancy-from cyclical downturns in the economy to the total disruption of nuclear war-has a major impact upon pension plans. The pension issues to be addressed in the years ahead are considerably broader than those for which solutions have already been reached. Chapter 7 ERISA: IS IT CONSISTENT WITH OTHER NATIONAL GOALS? (Prepared by Alicia H. Munnell,* Vice President and Economist, Federal Reserve Bank of Boston, MA) A. INTRODUCTION The Employee Retirement Income Security Act of 1974 [ERISA] was enacted in response to documented weaknesses of the private pension system. Prior to the Federal legislation, some employers imposed such stringent vesting and participation standards that many of their workers reached retirement age only to discover that because of some layoff, merger, or other event causing a break in service, they were not eligible for a pension. Even workers who sat- isfied their plans' participation and vesting requirements had no assurance that accumulated pension fund assets would be adequate to finance benefits. Although employers were expected to fund their plan over periods of 10 to 40 years, they were not legally re- quired to do so. Workers covered by inadequately funded plans risked losing pension benefits if their plans were terminated. Not only were many plans funded inadequately, but also a few were administered in a dishonest, incompetent, or irresponsible way. Other forms of financial manipulation, while not illegal, also jeopardized the welfare of plan participants. Pension assets were often concentrated in the stock of the plan-sponsoring company or used to make large loans to the company. This disregard for diver- sification often left workers' benefits dependent on the vagaries of the local economy or the financial condition of the firm. The outstanding characteristic of the pre-ERISA era was that pension plan participants were generally at the mercy of plan sponsors. If the pension plan were insufficiently funded or suffered investment losses, the claimants could conceivably forfeit part or all of their benefits. This risk was not consistent with the economic reality that pension benefits were often negotiated as forms of de- ferred compensation to which workers were legally entitled.' Pen- sion reform legislation was needed to redefine the rights of workers and to secure their benefit claims.

*The views expressed are those of the author and do not necessarily reflect official positions of the Federal Reserve Bank of Boston or the Federal Reserve System. I For evidence that workers trade off current wages for future retirement benefits see Bradley Schiller and Randall Weiss "Pensions and Wages: A Test for Equalizing Differences," Review of Economics and Statistics, vol. 62 (November 1980), pp. 529-38; and Ronald Ehrenberg and Robert Smith, "The Wage/Pension Trade-Off," in President's Commission on Pension Policy, Coming of Age, pp. 1200-18. (172) After 10 years of hearings and prolonged debate, Congress adopt- ed ERISA in 1974. The legislation's principle objective was to secure the rights of plan participants so that a greater number of covered workers would receive their promised benefits. To this end, regulations were promulgated in five areas of pensions: reporting and disclosure of plan administration; employee participation and vesting standards; funding schedules and fiduciary integrity; retire- ment plans for the self-employed; and the delivery of vested bene- fits. ERISA's participation and vesting standards enable workers to establish a legal claim to benefits. The implementation of funding and fiduciary standards and the establishment of the Pension Ben- efit Guaranty Corporation [PBGC] ensure that money will be avail- able to pay these legal benefit claims. Most observers agree that ERISA has been successful in meeting its stated objectives. The purpose of this chapter, however, is to assess the extent to which the provisions and implications of ERISA are consistent with other Government policies, namely, Federal tax provisions, employment policy, retirement income ob- jectives, and capital formation goals. Three general conclusions emerge from the following analysis: First, any consistency between ERISA and other Government policies is purely coincidental, since almost no consideration was given to the impact of ERISA on employment, overall retirement income, or capital formation during the legislative process. Even consistency with tax policy, which was cited as a motivation for Federal regulation of private plans, was not given much weight during the debates. Second, in some areas, such as employment, no clearly defined and internally consistent national policy exists, so that any assess- ment of the consistency of ERISA with, say, employment objectives is necessarily ambiguous. Finally, to the extent it is possible to discern national objectives in tax, employment, retirement income, or capital formation policy, the provisions and implications of ERISA appear more or less con- sistent with these goals, although some interesting anomalies do exist. Two caveats are in order before the analysis begins: First, although ERISA explictly sanctions defined contribution plans as a legitimate form of retirement savings, this chapter fo- cuses almost exclusively on defined benefit plans. The reason is that the main thrust of ERISA was to change the basic provisions of defined benefit plans, not to modify the nature of defined contri- bution plans. The exception was the introduction of IRA's and the expansion of Keogh plans, both of which are included in the discus- sion. Hence, whereas a study of the interaction of all pension plan provisions with national goals would necessarily include an analy- sis of both defined benefit and defined contribution plans, a consid- eration of the consistency of ERISA's goals with other Federal poli- cies seems appropriately limited to defined benefit plans. Second, by the nature of the assignment, this chapter differs from the earlier chapters in this volume. The previous authors have addressed historical issues where survey data, personal recol- lection, or tightly reasoned arguments can provide relatively pre- cise answers to the questions raised. This chapter, being so broad in coverage, is by necessity somewhat more speculative and open- ended. Hence, the goal is not so much to provide the definitive analysis of each issue, but to construct a bridge between the retro- spective assessments in this print and the future policy debates. B. ERISA AND FEDERAL TAX POLICY A 1965 report by the President's Committee on Corporate Pen- sion Funds cited the favorable treatment of private pensions under the tax laws as a reason for these plans to be the focus of public policy. 2 Under current law, compensation in the form of employer contributions to qualified pension and profit-sharing plans is de- ductible, like wages, by the employer when the contributions are made. Unlike wages, however, it is not taxed to the employee until the benefits are distributed to beneficiaries. By allowing the deferral of taxes until retirement, compensation in the form of pension contributions offers three advantages over compensation in the form of wages: First, the full dollar of contribution, without any reduction for income tax, is available for investment during the employee's working years-in sharp contrast to the situation in which a dollar is paid in wages and the employee has only the after-tax dollar to invest. Second, no tax is currently paid on the investment income from accumulated pension assets, whereas interest earned by the em- ployee on his ordinary savings is subject to tax as income accrues. Finally, when benefits are distributed, they are likely to be taxed at a lower marginal rate than if they had been taxed as they ac- crued to the employee, since the employee's income is usually lower after retirement and subject to the double exemption for per- sons over 65. The special tax provisions for pension plans date from 1921, but it was not until 1938 that Congress attempted through the tax code to influence the design and operation of private pension plans. To prevent pensions from being used simply as tax avoidance schemes, the Revenue Act of 1938 included a "nondiversion rule" to make pension trusts irrevocable, so that employers could not take large deductions for contributions during years of high earnings and then recapture those earnings by revoking the trust in poor years. To insure that pensions were not established solely for small groups of officers and key employees in high income brackets, the Revenue Act of 1942 completely revised the definition of an exempt pension or profit-sharing trust and significantly broadened the par- ticipation standards by requiring that plans be nondiscriminatory to qualify for favorable tax status.3 Numerous regulations were issued over the next 12 years to insure that the tax incentives aimed at encouraging the growth of

2 President's Committee on Corporate Pension Funds and Other Private Retirement and Wel- fare Programs, Public Policy and Private Pension Programs: A Report to the President on Pri- vate Employee Retirement Plans (GPO, 1965), p. 11. 3 In addition to increasing coverage by broadening participation standards, this act contribut- ed to the growth of pensions by allowing companies to receive assurance from the Internal Reve- nue Service before instituting a plan that employer contributions would be tax deductible. For further discussion, see Alicia H. Munnell, The Economics of Private Pensions (The Brookings Institution, 1982), p. 10. pension plans were not abused. These regulations were designed to prevent discrimination in favor of shareholders, officers, supervi- sors and other highly compensated individuals in either coverage, benefits, or financing. Although the code was extensively revised in 1954, few substantive changes were made in the provisions pertain- ing to private pension plans.4 Thus, the goal of Federal tax policy since 1942 has been to en- courage, through favorable tax provisions, the use of tax qualified pension and profit-sharing plans to insure greater retirement secu- rity for employees in general and not just the highly paid few. In other words, the rationale for the favorable treatment of qualified plans seems to be to secure retirement benefits for rank and file workers by providing tax incentives that will induce higher paid employees to support the establishment of employer-sponsored pen- sion plans. Before ERISA was enacted, however, the code contained no mandate for the IRS to be concerned about the actuarial sound- ness of pension plans nor did it provide much protection for the pension rights of individual participants. Title II of ERISA, which is, in turn, an amendment to the Inter- nal Revenue Code of 1954, established participation, vesting, and funding standards that must be satisfied for a plan to qualify for favorable tax treatment. Title II also contains provisions designed to expand coverage of tax-favored forms of retirement saving and provisions that limit the benefits or contributions that can be pro- vided to any one individual under a qualified plan.5 Since the major thrust of ERISA was to insure that employees in firms with qualified plans actually end up receiving benefits, the legislation's provisions are generally consistent with established federal tax policy. The participation and vesting standards of ERISA, however, go beyond the traditional discrimination concerns and are aimed at ensuring the accrual and preservation of benefits. These provisions reflect the conviction that favorable tax treat- ment can be justified only if it results in the actual receipt of sig- nificant retirement income for broad groups of employees.

1. ENFORCING BROAD-BASED RECEIPT OF BENEFITS Title I of ERISA embodies a series of minimum standards de- signed to enhance the likelihood that workers will secure pension rights.6 Vesting and participation rules brought workers who might otherwise have been excluded into pension plans and helped ensure that long years of service would result in eventual pension receipt. To prevent the violation of the spirit of earlier vesting by employ- ers "backloading" the benefit formulas, ERISA introduced rules de- signed to prevent the disproportionate accrual of benefits until late in the worker's employment. It also established specific guidelines

4 See Dan M. McGill, Fundamentals of Private Pensions, 5th ed. (Richard D. Irwin, Inc., 1984), p. 30. * Under ERISA, the largest annual benefit payable to any individual under a defined benefit plan was the lesser of $75,000 or 100 percent of average annual compensation during his three consecutive highest paid years. Under a defined contribution plan, the annual addition to an employee's account was hmited to the lesser of $25,000 or 25 percent of compensation. These limits were subject to the automatic cost-of-living adjustments applied to social security benefits. In 1982, the Tax Equity and Fiscal Responsibility Act reduced these limits from their adjusted levels to $90,000 for defined benefit plans and $30,000 for defined contribution plans. 'The minimum standards are discussed in greater detail in chs. 1, 2, and 6. for defining a "break in service." The combined impact of the par- ticipation, vesting, backloading, and break-in-service rules has im- proved employees' chances of accruing the right to pension bene- fits. To ensure that vested benefits are actually paid, ERISA estab- lished funding schedules, mandated fiduciary standards, and cre- ated the Pension Benefit Guaranty Corporation.' It requires that normal costs be fully funded each year, that liabilities from em- ployees' prior service be amortized over not more than 40 years, and that new liabilities arising from either the formation of new plans or benefit liberalizations in old plans be amortized over a 30- year period. ERISA also established fiduciary responsibility for pension plan trustees, for investment managers, and for any other person who may have control over the assets of the plan. Responsi- ble financial management combined with stricter funding stand- ards should ensure that adequate funds are available to pay bene- fits. To protect participants against the possibility that their plans might terminate with insufficient assets to cover benefit commit- ments, ERISA established the PBGC. This agency guarantees pay- ment of what are essentially vested benefits, up to a specified dollar limit, in the event of plan termination.' In short, ERISA's provisions make it much more likely that an employee covered by a qualified pension plan will receive a benefit in retirement. This development is fully consistent with the goal of Federal tax policy-namely, that rank and file workers receive sup- plemental retirement income by providing tax incentives that en- courage the establishment of pension plans. 2. IRA's-THE EXCEPTION Although the general thrust of ERISA is consistent with national tax policy, the provision included in ERISA to expand the coverage of private pension arrangements has resulted in a situation that di- rectly contradicts the code's nondiscrimination criterion. In an attempt to offer retirement income opportunities more widely, ERISA authorized a new form of retirement plan, namely the individual retirement account. Beginning in 1975 persons not covered by either an employer plan or a Keogh plan could set up an IRA and make tax-deductible contributions equal to 15 percent of their income, up to a maximum of $1,500. Almost immediately, however, an equity question arose. Many employees covered by pension plans were denied tax-deductible IRA contributions even though their employer's contribution to the plan on their behalf might be quite small or they might never vest in a retirement ben- efit because of frequent job changes. In some instances employees had withdrawn from active plan participation in order to set up

' Funding requirements, fiduciary standards, and benefit guarantees are discussed in chs. 1, 2, and 6. 'The PBGC guarantees the payent of what it terms "basic benefits," which include all re- tirement, death, and disability benefits of current retirees and their beneficiaries, and the normal retirement benefit payable in monthly installments for vested current employees. Basic benefits do not include lump-sum and special supplementary benefits payable under some plans to encourage early retirement, or death and disability benefits not in current payment status. See Alicia H. Munnell, "Guaranteeing Private Pension Benefits: A Potentially Expensive Busi- ness," New England Economic Review (March/April 1982), pp. 24-47. their own IRA's.9 In an attempt to alleviate this inequity, the Eco- nomic Recovery Tax Act of 1981 expanded eligibility to encompass all workers, including those currently covered by pension plans. While the introduction of IRA's and subsequent expansion of eli- gibility was consistent with the Nation's policy of providing tax in- centives to encourage saving for retirement, it did not reflect the second requirement that favorable tax provisions be enjoyed on a nondiscriminatory basis. Without this stipulation, the incidence of utilization of the IRA deduction is significantly greater for higher paid than for lower paid employees. Indeed, survey data indicate that 58 percent of individuals earning over $50,000 contributed to an IRA in 1982 compared to only 17 percent of people earning be- tween $15,000 and $20,000 (see table 1). This disparity is signifi- cantly greater than that found in pension coverage.' 0 Thus, the IRA provisions introduced by ERISA violate the basic intent of tax policy in the pension area-namely, to encourage the establish- ment of pension provisions that ensure employees at all levels of compensation relatively equal retirement protection.

TABLE 1.-USE OF INDIVIDUAL RETIREMENT ACCOUNTS IN1982, BY 1983 EARNINGS LEVELS

1983 earningslevels Per with

Total...... 16.9 $1 to $4,999 ...... 7.1 $5,000 to $9,999 ...... 8.5 $10,000 to $14,999 ...... 11.1 $15,000 to $19,999 ...... 17.3 $20,000 to $24,999 ...... 20.1 $25,000 to $29,999 ...... 28.4 $30,000 to $49,999 ...... 38.7 $50,000 and over ...... 57.6

19 " Datafrom Employee Benefit Research Institute/Department of Healthand Human Services, Current Population Survey Pension Supplement (May 83krcentages excluderespondents whoseearnings were not reported. Seuce:Emloeenefit ResearchInstitute, "Individual Retirement Accounts: Characteristics andPolicy implications," EBRI Issue Brief No. 32 (July1904), tble 3, p.6.

3. MAINTAINING INTEGRATION-PERHAPS ANOTHER EXCEPTION? Although private pension plans must be nondiscriminatory to qualify for favorable tax treatment, an employer is permitted to consider its contributions to Social Security in determining wheth- er its plan discriminates in favor of officers, shareholders, or higher paid executives. This approach is based on the premise that public and private retirement programs should function as a uni- fied system. Thus, if the Social Security program is weighted in favor of low-income workers, the private pension system can favor high-income workers so that, when coordinated with Social Securi- 9 See Statement of Daniel I. Halperin, Deputy Assistant Secretary for Tax Policy, before the Subcommittee on Oversight, Committee on Ways and Means (Oct. 5, 1978), in Department of the Treasury, News (October 1978). ao According to the same survey data, 74 percent of individuals earning over $50,000 were cov- ered by employer-sponsored pension plans in 1982, compared to 64 percent of people earning between $15,000 and $20,000. See Employee Benefit Research Institute, "Individual Retirement Accounts: Characteristics and Policy Implications," EBRI Issue Brief No. 32 (July 1984).

37-569 0 - 84 - 12 178 ty, all workers will have the same percentage of their earnings re- placed when they retire. Congress incorporated this reasoning into the tax code in 1942 by including the proviso that public and pri- vate retirement benefits would be considered together in determin- ing whether a plan was discriminatory." Basically, as long as the ratio of combined benefits to earnings is no higher for employees whose wages exceed the taxable wage base than for those whose wages are fully covered, a plan is held to be nondiscriminatory. The problem with the integration guidelines is that they allow or encourage the payment of tax-subsidized benefits to higher paid workers, while permitting the denial of benefits to some lower paid ones. For example, some integration formulas seem to be based on the assumption that Social Securuity almost fully replaces the pre- retirement earnings of low- and middle-income workers. 12 Yet, this is not the case. As shown in table 2, Social Security benefits amount to only 35 percent to 47 percent of preretirement earnings for individuals and couples in the bottom two quintiles of the income distribution. Even taking into account the nontaxable qual- ity of Social Security benefits for this group, replacement rates still do not exceed 65 percent of their earnings before retirement. This means that lower paid employees and their families will experience a substantial drop in their standard of living in retirement, unless they receive supplementary private pensions or have saved consid- erable sums on their own.

TABLE 2.-MEDIAN SOCIAL SECURITY REPLACEMENT RATES-BEFORE AND AFTER TAX-FOR MEN, NONMARRIED WOMEN, AND MARRIED COUPLES, BY PRERETIREMENT EARNINGS LEVELS x

MedianSocial Security replacement rate Medianafter-tax Social Security replacementrate Preretirementearnings quintiles m Nonmarried Married Nunarried Almen women couples Allmen womre couprMarried

Lowest ...... 44 47 na na 60 63 Second...... 35 39 ...... 47 51 Third...... 32 33 ...... 44 44 Fourth...... 26 28 ...... 42 36 Highest...... 18 3 22 ...... 33 28

NA-not available. SocialSecuirty benefits as a percentof estimatedtotal price-indexed earnings in the highest 3 of thelast 10 years. 'Quintilesderived from combined earnings distribution of all samplerespondents. Quintile boundaries are-$6,283, $9,290,$12,780, and $16,246. * Basedon 50 orfewer cases; subject to highsampling variability. Source:Alan Fox, "Earnings Replacement Rates and Total Income: Findings from the Retirement History Study" Social Security Bulletin vol. 45, No.10 (October1982), tables 7 and10, pp. 13 and 20.

" For a thorough discussion of the issues surrounding integration, see James H. Schultz and Thomas D. Leavitt, Pension Integration: Concepts, Issues and Proposals, Employee Benefit Re- search Institute (Washington, DC, 1983) and Raymond Schmitt, "Integration of Private Pension Plans With Social Security," Studies in Public Welfare, paper 18, prepared for the Subcommit- tee on Fiscal Policy of the Joint Economic Committee, 93 Cong. 2 sess. (GPO, 1974), pp. 174-75. 12 The current guidelines include provision for integration through either an 'excess" or "offset" method. nder a pure excess plan, the employer can limit benefits to that portion of an employee's final average pay in excess of the "integration level." The integration level is a career average of the social security wage base applicable to a particular employee. For in- stance, an employee reaching retirement age in 1984 would face an integration level of $12,840, and an integrated plan could pay benefits in 1984 up to 37.5 percent of final pay in excess of $12,840 without providing any benefits below that level. Under the offset method of integration, an employer can reduce a worker's pension benefit by as much as 83Vs percent of a worker's primary social security benefit. 179 During its deliberations on ERISA, Congress recognized that in- equities existed under the IRS integration guidelines. Subsequent- ly, both the House Committee on Ways and Means and the confer- ence committee between the House and Senate voted a freeze on further integration pending full consideration of the issue. In the face of last minute opposition by some employers and pension prac- titioners, however, the freeze was canceled.' 3 The question remains whether the provisions of ERISA that permit the continuation of existing integration procedures are consistent with the goals of Federal tax policy. Supporters defend integration on the ground that without it em- ployers would have to provide more than 100 percent of preretire- ment pay to lower paid workers in order to ensure adequate bene- fits (as a percent of pay) to higher paid ones. They contend that the cost of these excessively high benefits for lower paid workers would result in much lower wages for these workers and an unnecessary reduction in their preretirement standard of living. Furthermore, they point out that the cost of reducing the permitted integration differentials would be substantial. Critics have repeatedly proposed reforms aimed at narrowing the potential discrepancy between benefits awarded to higher paid as compared to lower paid workers under various integration meth- ods. These reform proposals are a response to the fact that the ex- isting integration procedures raise some troubling basic issues. If workers in fact pay for pension benefits throughout their working lives through reduced wages, it seems difficult to justify provisions that can deny lower paid employees any benefits.

4. THE IMPLICATIONS OF FAVORING EMPLOYER CONTRIBUTIONS ERISA maintained the existing asymmetry between the tax treatment of employer and employee contributions to qualified pen- sion plans. While the tax on employer contributions is deferred until benefits are paid in retirement, mandatory employee contri- butions are made from net after tax income. As a result of this asymmetry, private pension plans in this country are financed almost entirely on a noncontributory basis.' 4 This emphasis on em- ployer-financed plans tends to work against the thrust of the tax code and ERISA of ensuring broad-based receipt of benefits. When pensions are financed by employers, employees have no immediate claim on contributions, even though they have implicit- ly accepted lower wages in return for pension coverage. Rather, en- titlement to benefits based on employer contributions hinges on specific provisions of the pension plan. Even though employees' rights were significantly improved by the participation and vesting standards established under ERISA, many workers covered by pri- vate plans who fail to complete the 10-year vesting requirement will not receive benefits.

13 See Ray Schmitt, "Major Issues Facing the Private Pension System," Congressional Re- search Service (GPO, 1978), pp. 16-19. "Employer contributions have increased from 58 percent of total private pension contribu- tions in 1940 to 94 percent in 1980. See Munnell, The Economics of Private Pensions, table 2.1, p. 11. 180 If employees were to finance their pensions through tax deducti- ble contributions, the tradeoff between pension coverage and wages would be made explicit. The employee would also have an immedi- ate right to the accumulated contributions even if ultimate entitle- ment to benefits remained contingent on satisfying a 10-year vest- ing standard. Accumulated contributions could then be transferred either to an IRA or to a new plan when the employee changed jobs. Thus, employee-financed plans would reduce the forfeiture of earned pension rights and would probably lead to greater portabil- ity within the private pension system. If plans were financed by employees, any inequities inherent in the current IRS integration guidelines would also become obvious. If low-paid workers, like their higher-paid counterparts, were to re- ceive lower wages in exchange for deferred pension income, there would be no justification for denying them benefits at retirement. Employee contributions would clarify the issue; either low-paid workers would contribute and become eligible for benefits, or they would make no contributions and rely solely on Social Security for retirement income. 5. SUMMARY The favorable provisions embedded in the tax code and restated in ERISA have been successful in encouraging the growth of pri- vate pension plans. Despite nondiscrimination requirements, how- ever, the tax code alone was not adequate to assure broad-based benefit receipt. ERISA's vesting, participation, and funding provi- sions and the institution of the PBGC have improved this situation. More people covered by pension plans will receive benefits now than would have before the passage of ERISA. The introduction of IRAs, the retention of existing integration rules, and the reliance on an employer-financed system, however, continue to work against the main thrust of ERISA. One issue remains which, though outside the scope of this chap- ter, is receiving increased attention. The tax-deferred status of pen- sion plans represents a loss to the Treasury of significant revenues. Although the precise estimate of this revenue loss is subject to con- troversy, the amount is undoubtedly quite large.15 The Treasury calculates a tax expenditure for public and private pension plans that amounts to $50 billion in 1984.16 In addition, the favorable tax treatment of IRA's is estimated to cost the Treasury another $9 bil- lion. While only half of the work force is covered by a pension and only 17 percent utilize IRA's, all taxpayers must pay higher taxes to make up for these foregone revenues. Critics of the current tax treatment of private pensions, identify- ing what they believe is an inequitable distribution of tax benefits, have frequently proposed either eliminating the favorable tax pro- visions for private plans, or making pension coverage universal so that all workers would enjoy the advantages of tax deferral. While

15 See Employee Benefit Research Institute, "Retirement Program Tax Expenditures: A Case of Unsubstantiated, Undocumented, Arbitrary Numbers," EBRI Issue Brief (April 1983), and Alicia H. Munnell, "The Erosion of the Tax Base by Employee Benefit Plans," Statement before Committee on Ways and Means, U.S. House of Representatives (February 1984), pp. A-1-A-5. 16 Special Analyses, Budget of the United States Government, 1985 (GPO 1984), p. G-47. ERISA's tenth anniversary will be marked by an evaluation of the policies embedded in its explicit goals, considerable attention will also be focused on the long-standing tax policy on which ERISA was built. C. ERISA AND EMPLOYMENT POLICY President Kennedy in 1962 directed the Committee on Corporate Pension Funds to explore, among other issues, how pension plans could "contribute more effectively to efficient manpower utilization and mobility." 17 In response, the committee reported on three sep- arate but interrelated effects of private pension plans on manpow- er-namely, their impact on (1) labor mobility, (2) employment op- portunity for older workers, and (3) retirement behavior. Underly- ing the committee's analysis of each of these aspects of manpower policy was an explicit or implicit statement of a desirable national objective. In some areas, these national goals are the same today as they were two decades ago; in other areas, some shifts have oc- curred. The effect of pensions on labor mobility has remained an impor- tant concern of manpower policy. As the committee's report noted: In a changing and growing economy, continuing shifts in the occupational, industrial, and geographical needs for manpower take place. Private pensions, along with seniori- ty and other benefits based on length of service, tend to reduce labor mobility by tying workers to a particular em- ployer.' 8 Maintaining a mobile labor force is considered just as desirable today as it was in the early 1960's.19 An efficient allocation of re- sources requires an adaptable work force willing to relocate in re- sponse to changing demands. Ironically, of course, any attempt to alleviate the adverse impact of pensions on labor mobility conflicts with one of the major motivations for the development of private pensions-namely, increasing job tenure. Excess turnover can also result in the costly and inefficient utilization of resources. The employment of older workers has become, at least theoreti- cally, a more important manpower goal than it was in the early 1960's. Despite the continuing liberalization of early-retirement provisions in private plans and the trend toward earlier retire- ment, national legislation seems premised on the desirability of keeping older workers in the labor force. The two most obvious recent developments are the 1978 amendments to the Age Discrimi- nation in Employment Act, which raise the age for mandatory re- tirement from 65 to 70, and the 1983 amendments to the Social Se- curity Act, which extend gradually the age for normal retirement benefits from 65 to 67, beginning after the turn of the century. As in the case of mobility, however, any changes in the structure of pension plans designed to encourage the employment of older

"' President's Committee on Corporation Pension Funds, Public Policy and Private Pension Programs, p. in. 18 Ibid., p. 27. 19 Susan Meredith Phillips and Linda Pickthorne Fletcher, "The Future of the Portable Pen- sion Concept," Industrial and Labor Relations Review (January 1977), pp. 197-204. workers directly contradict one of the primary motives for the in- troduction of pensions-namely, retiring superannuated employees. With regard to the third dimension of manpower policy, retire- ment behavior, the committee argued that provisions should be completely neutral. That is, older employees should not be forced out of the labor force to alleviate unemployment problems for younger cohorts, yet adequate retirement income ought to be avail- able for those who choose to retire. 20 Today, national policy is schizophrenic on the issue of retirement. Pension arrangements in the private sector are designed to enable earlier and earlier retire- ment, while public policy developments are generally aimed at en- couraging older workers to remain in the labor force. For example, the 1983 Social Security Amendments increased the delayed retire- ment credit for workers who postpone retirement beyond the normal retirement age and the provision that gradually extends the retirement age will further reduce the benefits available to workers who choose to retire early. Having identified some of the major goals of employment policy that might be affected by private pension plan regulations, the question is now whether the reforms enacted in ERISA are gener- ally consistent or inconsistent with these goals.

1. THE IMPACT OF ERISA ON MOBILITY The main characteristics of pension plans that impede labor mo- bility are delayed vesting and the lack of post-termination index- ing. The absence of portability is also discussed in this context, but it is unclear how improved portability would reduce the potential loss for the mobile employee unless accompanied by some form of post-termination indexing. ERISA probably improved labor mobili- ty by accelerating vesting, but significant impediments remain since the legislation failed to address the impact of inflation on the vested benefits of terminated employees.

(A) VESTING The adverse impact of delayed vesting on labor mobility is straight forward. A worker who leaves coverage of a pension plan before attaining the age and service requirements for vesting loses all claim to the potential benefits. 2 1 The importance of this factor in a job change decision depends on the level of promised benefits, the proximity of the qualifying date, and the age of the worker. The hypothesis that pension plans influence labor mobility has been confirmed in several studies that compare the behavior of workers covered by a pension plan with that of noncovered work- ers. 22 A more recent and ambitious study by Bradley R. Schiller

20 President's Committee on Corporate Pension Fund, Public Policy and Private Pension Pro- grams,21 pp. 31-32. Robert Taggart, "The Labor Market Impacts of the Private Retirement System," Studies in Public Welfare, paper 11, prepared for the Subcommittee on Fiscal Policy of the Joint Economic Committee, 93 Cong. 1 sess. (GPO, 1973), p. 60. 22 Herbert Parnes, et al., The Pre-Retirement Years, vol. 4 (Columbus, Ohio: Center for Human Resource Research, Dec. 1974); Vincent S. Flowers and Charles L. Hughes, "Why Em- ployees Stay," Harvard Business Review 51 (July-August 1973), pp. 49-60; Richard B. Freeman, 'Non-Wage Effects of Trade Unions on the Labor Market: An 'exit-voice' Model," (Harvard Uni- versity (processed February 1976)); and Robert Taggart, "The Labor Market Impacts of the Pri- Continued and Randall D. Weiss attempted to move beyond the pension-no pension dichotomy and determine to what extent the specific fea- tures of a pension plan affect firm attachment. 23 Using characteris- tics for 133 of the largest plans and earnings records for the work- ers in the firms sponsoring these plans, Schiller and Weiss tested the impact of particular vesting and benefit provisions on the prob- ability of quitting. Their results with regard to the impact of vest- ing on firm attachment were as predicted: (1) Quit rates fell among young workers as they approached vesting, and then rose among vested workers once their benefits were secured, and (2) stringent vesting requirements significantly increased quits among younger workers, since the discounted expected value of benefits is very small for workers who must wait many years to become vested. Al- though the results were consistent with economic theory, their overall explanatory power was quite low. Nevertheless, these re- sults offer some insight into what impact the vesting changes in ERISA have had on labor mobility. ERISA undoubtedly has resulted in more rapid vesting. The law specifies three alternative vesting procedures, but most U.S. compa- nies have adopted the option that provides 100 percent vesting after 10 years of service. 2 4 Whereas prior to ERISA some plans re- quired employees to remain until retirement in order to receive a benefit, now almost any worker who leaves a plan after 10 years can be assured a retirement benefit. Accelerating vesting has probably reduced somewhat the quit rate among younger workers. Moving up the date at which benefits are vested greatly increases the discounted expected value of bene- fits for this group and lowers the likelihood that these workers will leave the firm. On the other hand, 10-year vesting has probably in- creased the likelihood of quits among workers with more than 10 years of service. Whereas previously these workers might have re- mained with the same employer 15 or 20 years to become vested, they now are assured a benefit after 10 years even if they leave. On balance, ERISA's vesting provisions have probably increased mobility.

(B) EFFECT OF INFLATION ON VESTED BENEFITS While the vested worker has less to lose from changing jobs than the nonvested worker, he will still receive a lower benefit as a result of moving than he would have received from continuous cov- erage in a single plan. This difference arises because pension bene- fits are generally based on final earnings levels. Thus, the worker who remains with the plan will receive benefits related to earnings just before retirement, while the mobile employee's benefits will be based on his earnings at the time he terminated employment. The magnitude of this effect depends on the rate at which earn- ings rise over the course of the employee's work life. Table 3 shows vate Retirement System," cited by Bradley R. Schiller and Randall D. Weiss, "The Impact of Private Pensions on Firm Attachment," The Review of Economics and Statistics (August 1979), pp. 369-380. "Schiller and Weiss, "The Impact of Private Pensions on Firm Attachment," p. 369. 24 Wliam J. Mischo, Sook-Kuen Chang, and Eugene P. Kaston, Corporate Pension Plan Study: A Guide for the 1980's, Bankers Trust Co., Employee Benefits Division (New York: Bank- ers Trust, 1980), pp. 338-40. 184 the ratio of benefits for workers with continuous and discontinuous employment under various assumptions about the rate of inflation and wage growth. Even though all employers on this example have identical plans, it is clear that the mobile employee is severely pe- nalized. If inflation were 6 percent, the pension received by the worker who had held four jobs during his work life would equal 51 percent of the pension received by the worker who had been con- tinuously employed by one firm. If inflation were 10 percent, the relative position of the mobile employee would deteriorate further, so that his benefit would be worth only 40 percent of that awarded to the one-job worker.

TABLE 3.-COMPARISON OF BENEFITS FOR A4-JOB WORKER AND A1-JOB WORKER

Compensation Ratioof benefits: Item baenaion rule(percent of Benefits 4-job/1job salary)3 worker

Inflation rate: 0 percent: 4-jobworker: 4 Job1...... $10,000 10 $1,000 ...... Job 2...... 10,000 10 1,000 ...... Job 3...... 10,000 10 1,000 ...... Job4 ...... 10,000 10 1,000 ...... Total...... 40 4,000 ...... 1.00 1-jobworker ...... 10,000 40 4,000 ...... Inflation rate:6 percent: 4-jobworker: Job1...... 17,908 10 1,790 ...... Job2...... 32,071 10 3,207 ...... Job3...... 57,435 10 5,744 ...... Job4...... 102,857 10 10,286 . Total...... 40 21,O27 ...... 51 1-jobworker ...... 102,857 40 41,143 ...... Inflation rate:8 percent: 4-jobworker: 4 Jobl ...... 21,589 10 2,159 ...... Job2 ...... 46,609 10 4,661 ...... Job 3 ...... 100,626 10 10,063 ...... Job4...... 217,243 10 21,724 Total...... 40 38,607 ...... '*... ' ...... 44 1-jobworker...... 217,243 40 86,897 ...... Inflation rate: 10percent: 4-job worker: 4 Job1...... 25,937 10 2,594 ...... Job2...... 67,275 10 6,728 ...... Job3...... 174,494 10 17,449 ...... Job4...... 452,593 10 45,259 ...... Total...... 40 72,030...... 4.0.7.2.,.O.3.0 ...... 40 1-jobworker...... 452,593 40 181,037 ......

'Assumesa consistentannual increase in wagesto compensatefor inflation,and no growthin wagesdue to productivity. 'Basesalary is $10,000and benefit iscalculated onearnings in last year of employment. 'Assumesannual benefit accrual of 1 percentayear. 'Assumesworker stays at each job for 10 years. SourceAlicia H. Munnel,The Economics of Private Pensions (the Brookings Institution, 1982), table 7-2, p. 176. 185 Thus the higher the rate of inflation, the more discontinuous em- ployment reduces the real value of benefits.2 5 This erosion of bene- fits occurs because benefits are not indexed between termination of employment and retirement. Post-termination indexing, however, meets with considerable re- sistance in the United States, even though such a proposal was re- cently introduced in Great Britain. 2 6 Employers are willing to pro- vide implicit indexation, by basing benefits on final salary, for people who remain covered by their plan until retirement, but they resist doing so for terminated employees. Part of the'reason might be that the firm may be forced to bear the full cost of increases due to unanticipated inflation for terminated employees, whereas in the case of active employees it can shift the burden by slowing the rate of wage growth. Furthermore, by providing lower benefits to mobile employees, the firm can perhaps reduce employee turnover and retain skilled workers. Without post-termination indexing, however, the mobile employee suffers a substantial loss in the value of his pension benefit because of inflation (see table 4). ERISA did not address this issue, and therefore, the potential ero- sion of vested pension benefits continues to provide a significant deterrent to labor mobility.

TABLE 4.-PURCHASING POWER OF $100 VESTED BENEFITS AT AGE 65, AT VARYING INFLATION RATES AND AGE AT JOB TERMINATION

Annualintftation rate Ageat jebterminationAnulifaonrt 6 percent 8 percent 10 percent

30 ...... $13 $7 $4 40 ...... 23 15 9 50 ...... 42 32 24 Source:Author's calculations.

(C) THE LIMITS TO THE POWER OF PORTABILITY Preserving the value of pension benefits between termination and retirement cannot be accomplished merely by increasing porta- bility. Literally, portability means nothing more than the ability for an employee to transfer the present monetary value of vested pension credits to a succeeding plan or a central clearinghouse 2 5Similar problems, of course, emerge under flat benefit formulas. A vested pension of $100 for an employee who terminates employment at age 30 would have the purchasing power of only $55 by age 65 if annual inflation were 4 percent; with an inflation rate of 8 percent, purchasing power would be reduced to $32. 2 See "Funds. No Place for Personal Views," Pensions and Investment Age (June 25, 1984), p. 10, and Department of Social Services, Personal Pensions: A Consultative Document (British Government Press, 1984). 186 upon termination of employment. 27 Some very minor changes were made in ERISA to facilitate this type of portability. 28 The transferring of vested pension credits to an IRA or central clearinghouse, however, does not prevent the erosion in the value of benefits as long as market interest rates do not exceed the inter- est assumption used by the actuary to calculate the discounted value of the future benefit. 2 9 Consider the example of a 50-year-old worker, who leaves a plan in which he has gained credits equal to 20 percent of final pay at age 65. At termination, his salary is $30,000, so he is entitled to a benefit of $6,000 per year beginning at 65. If the actuaries assume an interest rate of 8 percent in calcu- lating the present value of this sum, the worker will receive $1,890 to transfer to his IRA. If he earns 8 percent in his IRA, he will have $6,000 at retirement. At this time, however, his salary will, most likely, have increased to $60,000, so that his pension if he had stayed with the firm would have been $12,000, or twice the value of his deferred vested benefit. Thus, transferring money to an IRA does not help the value of the pension keep pace with salary growth. The only possible advantage to the employee lies in the chance that the interest rate in the market is higher than that as- sumed by the actuary. Such a discrepancy would be unlikely to per- sist, however, if portability provisions became widespread. Hence, increasing the portability of vested benefits would be unlikely to increase labor mobility. The concept of portability has gained substantial appeal, howev- er, because many people have incorrectly interpreted it as a means for allowing the mobile worker to count all service with all employ- ers toward meeting the conditions for a pension.3 0 This type of "portability" is found in collectively bargained, multiemployer plans, under which all of an employee's service for any of the con- tributing employers is treated as if it were service for a single em- ployer. This enables employees, who can expect to work for many different employers and in several geographical areas during the course of their working lives, to accumulate sufficient service to qualify for a benefit. Enhancing this type of transfer of service credits among defined benefit plans would reduce the loss suffered by a nonvested employee and would enhance labor mobility. This type of reform, however, was not considered in ERISA. In summary, then, ERISA's vesting provisions have probably re- duced mobility among younger workers, but facilitated mobility among workers with more than 10 years of service. For this latter

27 Dan M. McGill, Preservation of Pension Benefit Rights (Richard D. Irwin, Inc., 1972), p. 1. 28 First, ERISA requires each plan to provide an annual statement to the IRS regarding em- ployees who have terminated employment that year with a right to a deferred vested benefit. This information is passed on to the Social Security Administration which will inform employ- ees of their accumulated pensions credits when they apply for Social Security benefits. Second, ERISA introduced "tax-free rollovers" whereby participants can transfer money from a quali- fled pension plan to an IRA or another qualified plan within 60 days without paying taxes. Fi- nally, ERISA directs the PBGC to provide advice and assistance in establishing IRA s or trans- ferring funds. 29 Several proposals have been made over the years to create a central clearinghouse to permit workers to transfer accrued pension credits. See Merton C. Bernstein, The Future of Pri- vate Pensions (London: Free Press of Glencoe, 1964), pp. 270-96; President's Committee on Cor- porate Pension Funds; and Donald S. Grubbs, Jr., "Pension Portability," National Pension Poli- cies: Private Pension Plans, Hearings before the subcommittee on Retirement Income and Em- ployment of the House Select Committee on Aging, 95 Cong. 2 sess. (GPO, 1978), pp. 529-65. so Ray Schmitt, "Major Issues Facing the Private Pension System," p. CRS-27. group, however, the potential loss in the value of frozen vested ben- efits probably still acts as an important deterrent to job changes. Nevertheless, on balance, ERISA has improved mobility, which is consistent with our national goals, but contrary to employers' moti- vations for introducing pension plans. 2. ERISA AND THE EMPLOYMENT OF OLDER WORKERS The most important characteristic of defined benefit plans that might affect the employment of older workers is the fact that the cost of providing benefits increases significantly with the age of the employee. The reason is that the earlier the age of entry and the longer the period of participation, the more time for interest to be earned on the employer's contribution and the greater the likeli- hood the worker will leave before qualifying for benefits. Thus, since older employees are closer to retirement and, therefore, more likely to remain under the plan and to have fewer years to earn interest income on the funds in their pension accounts, the cost of offering pensions to this group is significantly higher than for younger workers. Since the magnitude of the discrepancy increases with the level of interest rates, the cost differential is significantly greater today than it was in the 1960's. In competitive labor markets, employers could pay older workers less in cash wages to compensate for the higher pension cost.3 1 If employers do not have control over their wage structure in the short-run, however, due to the existence of union contracts or other institutional arrangements, an alternative strategy for controlling costs is to encourage turnover among older employees, through varying working conditions and other nonwage characteristics of employment, and to avoid hiring older workers. ERISA contained two provisions that, at first glance, might be thought to mitigate the adverse impact of pension costs on hiring and retaining older employees and one provision that would defi- nitely make the hiring of older workers less attractive. The two apparently favorable provisions are the ability to ex- clude from a pension plan any employee hired within 5 years of the normal retirement age and the cessation of benefit accruals for an employee after age 65.. In theory, both of these provisions should reduce the cost of hiring an older employee and thereby increase the likelihood of such an event. Setting the maximum participation age at 5 years before normal retirement, however, represents an in- crease over earlier practices, under which the maximum age was often 10 to 15 years before retirement. 32 In addition, it seems un- likely that many firms would consciously employ older workers to avoid pension costs, since this would run counter to the purpose of having a plan in the first place. The second provision of not requir- ing any benefit accrual after the normal retirement age probably has little impact on the hiring of older workers and also appears grossly inconsistent with the 1978 legislation prohibiting mandato- ry retirement before age 70.

31 Burt S. Barrow and Ronald G. Ehrenberg, "The Costs of Defined Benefit Pension Plans and Firm Adjustments," Quarterly Journal of Economics, vol. XCIII (November 1979), pp. 523-40. 32 Taggart, "The Labor Market Impacts of the Private Retirement System," table 35, p. 102. Although two ERISA provisions are aimed at controlling the cost differential under defined benefit plans between old and young workers, one of the vesting provisions included in ERISA exacer- bates the problem. Companies that adopt the "rule of 45" vesting schedule, whereby 50 percent vesting occurs when the sum of the employee's age and service (a minimum of 5 years) equals 45 and 10 percent vesting occurs each year for the next 5 years, substan- tially increases the relative costs of hiring older workers. If a worker aged 40 were hired and worked for 5 years, he would be 50 percent vested at the end of the period. A 30-year-old worker would not have any vested rights after 5 years and a 25-year-old would have no vested rights until the 10th year of service. Thus, although two provisions contained in ERISA had some po- tential for stimulating the employment of older workers, neither standardizing the maximum participation age nor allowing benefit accrual to cease after age 65 has succeeded in relieving employers of the high costs of hiring and retaining these workers. This, in conjunction with the adverse effect of the "rule of 45" on the costs of hiring older workers, indicates that the passage of ERISA did not substantively reflect the national desire to increase employ- ment opportunities for older workers.

3. ERISA AND RETIREMENT BEHAVIOR One of the major reasons that pension plans were instituted was to provide employers with an impersonal and egalitarian means of retiring older workers in order to restructure the age composition of their work force.33 Recent studies have confirmed that economic factors, rather than health and other noneconomic considerations, are the primary determinents of the retirement decision, and that, all else equal, the higher the pension benefit that a worker can expect to receive in retirement, the more likely he is to withdraw from the labor force. 34 Any change in benefit levels or the age at which benefits become available should have an important impact on retirement behavior. While ERISA contained no special provisions aimed at the amount or timing of benefits to be provided by a qualified pension, the entire thrust of the legislation was to ensure that individuals covered by pension plans receive pension benefits. Since the result of the legislation has been that more people reaching retirement age will have supplementary private pension benefits than would have in the absence of ERISA, the net impact of ERISA has been to encourage earlier retirement. This effect has probably been mitigated somewhat by the fact that inflation has dominated much of the post-ERISA era and the legislation contained no provisions to ensure that plan sponsors offer some form of postretirement cost-of-living adjustment. With- out such adjustments, retirees' living standards will decline in re- tirement as inflation erodes the purchasing power of their benefits.

3 William Graebner, A History of Retirement: The Meaning and Function of an American Institution, 1885-1978 (Yale University Press, 1980), pp. 18-53. 4 The most careful analysis of the effect of pension benefits on the retirement decision is Gary S. Fields and Olivia S. Mitchell, Retirement, Pensions and Social Security (Cornell Univer- sity, 1984; forthcoming). For further studies, see bibliography included in the Fields-Mitchell monograph. Since the future pattern of price increases cannot be predicted, no amount of preplanning can insure protection from inflation. For private plans to provide benefits that keep pace with infla- tion, however, they must be able to earn a return in fund assets that fully reflects increases in the price level. In other words, pen- sion funds must have access to investments whose real rate of return is not affected by inflation. Most of the evidence in the eco- nomic literature, however, suggests that real returns on corporate equities and fixed-income securities decline in the face of unantici- pated increases in inflation.35 Hence, plan sponsors cannot offer fully indexed benefits without incurring higher real costs. On the other hand, although the returns on assets do not fully incorporate the effects of inflation, nominal yields do rise as the rate of price increases accelerates. Although many plan sponsors currently pro- vide some ad hoc cost of living increases, others appropriate the partial inflation premium incorporated in asset yields to reduce their costs, so that the real costs of providing pensions are lower than would be required in the absence of inflation. To increase the number of pension recipients eligible to receive COLA's, all plan sponsors could be required to value the firm's liability for benefits in payment status at a real rate of return of 2 or 3 percent and to use anything above this amount to provide partial post-retirement cost-of-living adjustments.3 6 This form of indexing will not impose any more real costs on plan sponsors than would be incurred in a noninflationary environment and will ensure beneficiaries roughly the same inflation protection as they could have secured on their own. Despite the lack of automatic postretirement cost-of-living adjust- ments, ERISA has probably led to more and earlier retirement through ensuring that plan participants actually receive pension benefits. This trend toward earlier retirement is not consistent with the public policy goal of keeping workers in the labor force for longer periods of time, and the increasing incentives in the private sector to retire early are going to create serious problems after the turn of the century when the baby boom reaches retirement age.

4. SUMMARY Since the United States does not have clearly defined employ- ment policies in the areas that might be affected by pension plan provisions, it is difficult to summarize the degree to which ERISA is consistent or inconsistent with national goals. The major impact of ERISA on manpower has probably been to induce more retire- ment by establishing provisions that ensure the receipt of pension benefits. This outcome is consistent with the interests of employers,

"See Franco Modigliani and Richard A. Cohn, "Inflation, Rational Valuation and the Market," Financial Analysts Journal, vol. 35 (March-April 1979), pp. 22-44; Bruno A. Oudet, "The Variation of the Return on Stocks in Periods of Inflation," Journal of Financial and Quan- titative Analysis, vol. 8 (March 1973, Papers and Proceedings, 1972), pp. 247-58; John Lintner, "Inflation and Security Returns," Journal of Finance, vol. 30 (May 1975), pp. 259-80 and Eugene F. Fama, "Stock Returns, Real Activity, Inflation and Money," working paper (University of Chicago,6 Graduate School of Business, 1979). 3 This type of indexation has been proposed in Canada. See Ontario Select Committee on Pen- sions, final report (Toronto, Government of Ontario, 1982). For a comprehensive discussion of indexation, see Robert J. Myers, Indexation of Pensions and Other Benefits (Richard D. Irwin, Inc., 1979). who maintain pension plans in large part to retire superannuated workers, but inconsistent with public policy interest in prolonging the work life and extending the retirement age. The other area where ERISA may have affected labor force activ- ity is through the impact of more rapid vesting on labor mobility. Since mobility is enhanced by the attainment of vested pension credits, the likelihood of quitting among workers with 10 years of service has probably increased. To substantially improve labor mo- bility, however, it would be necessary to institute post-termination indexing, since inflation rapidly erodes the value of vested benefits for the mobile employee. Like the retirement issue, however, the interests of the public policymakers and private employers diverge on the desirability of improved mobility. From a public perspective, national output will be increased by having workers move quickly to those areas where they can be most productive; from the per- spective of the individual employer, however, costs can be reduced and productivity enhanced by retaining highly trained workers. D. ERISA AND RETIREMENT INCOME POLICY National retirement income goals have been expressed in a vari- ety of ways-through Federal legislation, by White House Confer- ences on Aging and, most recently, by the President's Commission on Pension Policy. 3 7 The recommendation of the last was the most specific and the most ambitious-namely, to ensure the full re- placement of preretirement disposable income.38 The White House Conferences on Aging have emphasized also the need for establish- ing a minimum level of retirement income and for maintaining the adequacy of the total benefit throughout the retirement period. To meet these retirement goals, a three-tiered system of retire- ment income maintenance has developed in the United States. This system consists of welfare programs, such as (1) the supplemental security income [SSI] program, which provides a minimum guaran- teed income to the needy elderly; (2) compulsory public contribu- tory programs, such as old-age and survivors insurance [OASI]; and (3) private provisions for retirement through private pensions and individual savings. In view of the substitutability of the programs, the potential role for private pensions in the provision of retire- ment income is determined by the gap between the income require- ments of the elderly and the benefits provided by government pro- grams-primarily Social Security. The extent to which Social Security benefits meet the income goals of the elderly can be assessed against two alternative meas- ures: an absolute standard of living and the maintenance of prere- tirement living standards. For measuring against an absolute standard, budgets constructed by the Bureau of Labor Statistics used to be a useful benchmark. These budgets represented the cost of hypothetical lists of goods and services for retired couples at three relative standards of living and were updated annually on

37 For a summary of these goals, see Elizabeth L. Meier, Cynthia C. Dittmar and Barbara Boyle Torrey, Retirement Income Goals, working paper for the President's Commission on Pen- sion Policy (March 1980). 38 President's Commission on Pension Policy, Coming of Age: Toward A National Retirement Income Policy (GPO, 1981), p. 42. the basis of changes in the Consumer Price Index and in consumer expenditure patterns. Since publication of the BLS budget data was discontinued in 1982, updated budgets can be estimated only by in- creasing the 1981 level by the Consumer Price Index, making no adjustment for changes in consumption patterns.3 9 Using this ap- proach, the low, intermediate, and high budgets for a retired couple were estimated as $7,900, $11,200, and $16,500, respectively, for the autumn of 1983. Even though these budgets may be too high as a level for welfare support, the intermediate budget was designated by the 1981 White House Conference on Aging as the minimum standard for aged couples in the United States and may therefore be considered a reasonable goal for combined Social Secu- rity and private pension benefits for a worker with a history of av- erage earnings and steady employment. 4 0 As shown in table 5, Social Security benefits in 1984 for a retired couple with average earnings amount to about 87 percent of the in- termediate budget. For those with a history of earnings at the tax- able maximum, benefits exceed the intermediate budget by 13 per- cent; and benefits for low-wage workers equal 57 percent of the in- termediate budget. However, the diagonal elements of the table may be the most relevant, since it is difficult to argue that a low- wage couple earning $7,000 in 1983 should receive a Social Security benefit of $11,200 in order to attain the intermediate budget. The diagonal percentages indicate that Social Security benefits to cou- ples with low, average, and maximum earnings amount to approxi- mately 82 percent, 87 percent, and 77 percent of the low, interme- diate, and high budgets, respectively. 4'

TABLE 5.-SOCIAL SECURITY BENEFITS FOR ARETIRED COUPLE, JANUARY 1984, AS APERCENTAGE OF 3 BUDGET LEVELS 1

1983earnings of retired worker and spouse, both age 652 L(Low Id9 et Interm($dbete )neieitHiHifh budget

Low earnings ($7,000)...... 0.82 0.58 0.39 Averageearnings ($15,100)...... 1.24 .87 .59 Maximumearnings ($35,700) ...... 1.60 1.13 .77

Budgetsare projected to autumn 1983. The projections are based on a 6.2 percentincrease in 1982 and a 3.2-percentincrease in 1983. 2 Theannual social security benefit amounts payable to a workerand spouse retiring in 1984with 1983 earnings equal to thelow, average and maximumearnings levels are $6,498, $9,770 and $12,665, respectively. Source:Author's calculations based on Departmentof Labor, Bureau of LaborStatistics, "Three Budgets for a RetiredCouple, Autumn 1981," BLS News(July 30, 1982). Social Security benefits can also be evaluated in terms of prere- tirement living standards rather than against a monetary stand- ard, in which case replacement rates rather than benefit levels are the relevant measure of benefit adequacy. Once a minimum level of income support is assured, a replacement rate-the ratio of ben-

" The last data were for 1981. See Department of Labor, Bureau of Labor Statistics "Three Budgets for a Retired Couple, Autumn 1981," BLS News (July 30, 1982), p. 1. 40 1981 White House Conference on Aging, "Committee Recommendations from White House Conference on Aging" (Nov. 30-Dec. 31, 1981), recommendation 23, p. 3. 4 ' Although the benefits in table 5 include the 50-percent supplement for a nonworking spouse, the percentages would be identical for retired individuals, assuming that the cost of sup- porting a couple is 50 percent greater than it is for an individual. 192 efits to preretirement earnings-is actually a more appropriate cri- terion against which to assess wage-related benefit programs. Retirees require considerably less than 100 percent of their pre- retirement income to maintain their standard of living. First, whereas preretirement earnings are subject to the Federal income tax, the Social Security payroll tax, and State and municipal income taxes, a large portion of retirement income is not taxed.4 2 Second, work-related expenses, such as transportation, clothing, and meals purchased away from home, are reduced during retire- ment. Finally, expenditures will also be lower for services, such as cleaning and cooking, that were purchased while a person worked but that retirees can perform for themselves. 43 Owing to lower taxes, lower expenditures for household services, and reduced work expenses, retirees require approximately 64 to 81 percent of prere- tirement earnings to maintain their preretirement living standards (see table 6).44

2 4 Social security benefits, which constitute the bulk of retirement income, are not taxable for most retirees. Beginning in 1984, however, up to one-half of social security (and Railroad Retire- ment Tier 1) benefits received by taxpayers whose incomes exceed certain base amounts will be included in taxable income. The base amounts are $25,000 for a single taxpayer, $32,000 for mar- ried taxpayers filing jointly, and zero for married taxpayers filing separately. Income for pur- poses of calculating these base amounts includes adjusted gross income and one-half of social security and Railroad Retirement Tier 1 benefits. The amount of benefits included in taxable income is the lesser of one-half of benefits or one-half of the excess of the taxpayer's combined income (adjusted gross income plus one-half of benefits) over the base amount. For more detail see John A. Svahn and Mary Ross, "Social Security Amendments of 1983; Legislative History and Summary of Provisions," Social Security Bulletin, vol. 46, No. 7 (July 1983), p. 26. "Some would argue that the availability of medicare reduces out-of-pocket medical expenses for the elderly as well. Medicare, however, requires certain out-of-pocket expenditures. More- over, many retired workers have had health insurance during their working years, so that their pre- and post-retirement out-of-pocket medical expenses may remain fairly constant. "For earlier estimates of the percentage of preretirement earnings required to maintain pre- retirement living standards, see Peter Henle, "Recent Trends in Retirement Benefits Related to Earnings," Monthly Labor Review, vol. 95 (June 1972), p. 18; and Jane L. Ross, "Maintenance of Preretirement Standards of Living after Retirement," Technical Analysis Paper 10, Office of Income Security Policy, Office of the Assistant Secretary of Planning and Evaluation, Depart- ment of Health, Education, and Welfare (August 1976). TABLE 6.-RETIREMENT INCOME NEEDED TO MAINTAIN PRERETIREMENT STANDARD OF LIVING FOR PERSONS RETIRING INJANUARY 1984, SELECTED INCOME LEVELS

Preretirementtaxes Reductionsin expenses at retirement Postretirementtaxes Equivalentretirement income Disposabe Net Grosspreretirement income Stateand preretirementWork related Savingsand investments preretirement Staleand Federal' Ica 2 income expenses3 Amount Percent iroe Federalircome local2 Amount Ratio

Single person: $7,000...... $1,114 $116 $5,770 $346 0 0 $5,424 0 0 $5,424 0.77 $8,000...... 1,331 143 6,526 392 $98 1.5 6,030 0 0 6,030 .75 $10,000...... 1,795 203 8,002 480 240 3.0 7,282 0 0 7,282 .73 $15,000...... 3,108 379 11,513 691 691 6.0 10,131 $150 $27 10,308 .69 $20,000...... 4,716 608 14,676 881 1,321 9.0 12,474 420 76 12,970 .65 $30,000...... 6,156 746 23,098 1,386 2,772 12.0 18,940 1,328 239 20,507 .68 $50,000...... 12,236 1,747 36,017 2,161 5,403 15.0 28,453 4,581 825 33,859 .68 $70,000...... 19,145 2,990 47,865 2,872 8,616 18.0 36,377 7,260 1,307 44,944 .64 Marriedcouple (filing jointly): $7,000...... 898 77 6,025 362 0 0 5,663 0 0 5,663 .81 $8,000...... 1,104 102 6,794 408 102 1.5 6,284 0 0 6,284 .79 $10,000...... 1,538 156 8,306 498 249 3.0 7,559 0 0 7,559 .76 $15,000...... 2,685 302 12,013 721 721 6.0 10,571 0 0 10,658 .71 $20,000...... 3,951 470 15,579 935 1,402 9.0 13,242 0 0 13,242 .66 $30,000...... 5,062 549 24,389 1,463 2,927 12.0 19.999 613 110 30,722 .69 $50,000...... 10,023 1,348 38,629 2,318 5,794 15.0 30,517 3,558 640 34,715 .69 $70,000...... 15,899 2,406 51,695 3,102 9,305 18.0 39,288 5,957 1,072 46,317 .66 1984Federal income and Social Security taxes. a Basedon Stateand local income tax receipts, which were 18 percentof Federalincome tax receipts from 1973-83. Does not include property tax. 3Estimatedas6 percentat disposableincome. Postretirementtaxesare o incomein excessof SocialSecurity benefits for single persons with retirement income below $25,000 and married couples with retirement income below $32,000. For all others,postretirement taxesinclude Federal incometaxes levied on V of SocialSecurity benefits received. Source:Author's calculations based on President'sCommission on Pension Policy, An Interim Report (GPO, November 1980), pp. 11-12. 194 Social Security replacement rates for various types of hypotheti- cal beneficiaries retiring at the beginning of 1984 are shown in table 7. As illustrated, workers aged 65 earning $15,100 (the aver- age earnings level in 1983) received a benefit in early 1984 equal to 43 percent of preretirement earnings; to maintain his preretire- ment standard of living, he would need approximately 69 percent of prior earnings (table 6). The replacement rate for a couple with average earnings eligible for the 50 percent supplementary spouse's benefit is 64 percent as compared to the 71 percent that is required to maintain a preretirement standard of living.

TABLE 7.-REPLACEMENT RATES BY TYPE OF BENEFICIARY, EARLY.1984 [Inpercent]

Preretirementearnings Ageand type of beneficiary Low Averag Maximum ($7,000)2 ($15,10) ($35,700)4 Single worker: Age65...... 62.2 43.0 23.7 Age62...... 49.7 34.4 18.9 Worker age65, with spouse: Age65...... 93.3 64.5 35.5 Age62...... 85.5 59.2 32.5 Worker age62, with spouse: Age65...... 80.8 55.9 30.7 Age62...... 73.0 50.5 27.8 Widow age65, spouseretired at- Age65...... 62.2 43.0 23.7 Age62...... 51.5 35.7 19.6 1903earnings levels. Assumesan annual income slightly below the minimum wage during working years. 3 Assumesan annualincome approximately equalto theaverage of total wages in eachyear of worklife. 4 Assumesincome equal to themaximum taxable amount each year of worklife. Source:Author's calculations based on unpublisheddata from SocialSecurity Administration, Office of the Actuary,July 1984. These comparisons, however, probably overstate the role played by Social Security in providing retirement income for two reasons. First, using the analytical construct of a hypothetical individual with a history of average earnings exaggerates the size of Social Se- curity benefits. Even though the "average earner" designation im- plies that replacement rates are being measured for retiring work- ers in the approximate middle of the earnings distribution of all re- tiring workers, that is simply not the case.45 Rather, the concept of average earnings is based on a composite of earnings for all work- ers at all stages in their careers and does not reflect the fact that the earnings of retiring workers are usually higher than those of their younger counterparts. 46 For instance, the average earnings for men in 1983 was $19,175, whereas the average for full-time men in their mid-50's and early 60's was $27,084. In other words, the "average" preretirement man in 1983 was actually earning closer to the taxable maximum ($35,700 in 1983), and the "average" re-

1 The following discussion is based on work by James H. Schulz. For a statement of this argu- ment, see "Assessing the Adequacy of Pension Income," testimony before the President's Com- mission on Pension Policy (Jan. 11, 1980). 4 See Nancy D. Ruggles and Richard Ruggles, "The Anatomy of Earnings Behavior," in F. Thomas Juster, ed., the Distribution of Economic Well-Being, National Bureau of Economic Re- search, Studies in Income and Wealth, vol. 41 (Ballinger for NBER, 1977), pp. 115-58. placement rate should therefore not be viewed as 43 percent but rather as 30 to 35 percent of preretirement earnings. The second factor that tends to exaggerate the relative level of Social Security benefits is the assumption that people retire at 65. In fact, 64 percent of the total number of Social Security benefits in current-payment status in early 1984 were actuarially reduced to reflect retirement before age 65.4 For the "average" worker, re- tirement at age 62 reduces the replacement rates from 43 to 34 per- cent. Indeed, data on actual replacement rates for men in the middle quintile of the income distribution showed benefits equal to 32 percent of preretirement earnings (table 2). In short, convention- al replacement rate analyses exaggerate the extent to which Social Security benefits replace preretirement income and consequently underestimate the role remaining for private pensions. If full replacement of disposable income is the Nation's retire- ment income goal and Social Security meets roughly three-quarters of this goal for the low-income recipient, half of the middle-income worker, and a third for the person with a history of maximum earnings, then private pensions have an important supplementary role to play throughout the income distribution. The question is whether the provisions of ERISA enhance or restrict the ability of private pension plans to fulfill their required role.

1. ENHANCING THE LIKELIHOOD OF BENEFIT RECEIPT The main thrust of ERISA was to ensure that more employees covered by qualified pension plans receive benefits in retirement. The participation, vesting, and insurance provisions in the legisla- tion significantly increase the probability of a covered worker re- ceiving a benefit. Some have argued, however, that ERISA has shifted the incen- tive to establish defined benefit plans to defined contribution plans, and encouraged employers to terminate existing plans, thereby lowering the number of workers who might otherwise have expect- ed to receive pensions from a defined-benefit plan.48 Indeed, a rash of defined benefit plan terminations did occur following the enact- ment of ERISA. Three important points, however, should be kept in mind when assessing the impact of ERISA on the number of people who will ultimately receive benefits. First, to the extent that abuse existed prior to ERISA, any at- tempt to impose constraints on the operations of pension plans will encourage those employers who had no intention of fulfilling bene- fit commitments to terminate their pension plans. Very little re- tirement income protection is lost by this type of termination. Second, the passage of ERISA was followed by an extremely severe recession, which put economic pressure on many businesses to terminate their plans. In surveys conducted by the PBGC to de- termine the reasons for plan termination, either business failures or generally adverse business conditions were cited as the major reason for plan termination in the majority of cases, while ERISA was singled out as an important factor by less than 17 percent of

47 Social Security Bulletin, vol. 47, No. 3 (March 1984), table Q-2, p. 54. 48 See ch. 5. the terminating employers. 4 9 The importance of the business cycle on plan terminations is reinforced by the fact that terminations de- clined sharply after 1976 as the economy recovered and then rose again in 1982 and 1983 in the wake of the severe economic down- turn. Finally, although many plans have been terminated since the passage of ERISA, there are more defined benefit plans today than existed in 1974. The number of these plans has grown by 42 per- cent, from 232,838 in 1974 to 330,485 in 1983.50 This increase oc- curred over a period when the labor force grew by only 16 per- cent.51 In short, ERISA's provisions designed to ensure the delivery of pension benefits will result in more workers receiving pensions than would have in the absence of ERISA. This aspect of the legis- lation is fully consistent with the nation's retirement income policy, which requires supplementation of Social Security benefits by private pensions in order to avoid a substantial decline in eco- nomic wellbeing upon retirement.

2. INTRODUCTION OF JOINT AND SURVIVORS OPTION One of the major failures of this Nation's retirement income system is illustrated by the large number of elderly single women with incomes below the poverty line. 5 2 This situation is attributa- ble, in part, to the fact that before 1974 most pension plans made no provision for retirees' surviving spouses. The passage of ERISA reversed this trend by requiring that pension plans offer spousal protection through the provision of joint-and-survivor annuities. Such an annuity consists of a worker's pension payable over the life of the participant plus a survivor pension, payable over the life of the surviving spouse. The extra protection offered by the joint- and-survivor option is usually purchased through reductions in the retiree's own pension and is provided automatically unless rejected in writing by the retiree. The introduction of the joint-and-survivor option is consistent with the Nation's retirement income goals and should contribute to the well-being of elderly widows.

3. GAPS IN PROTECTION NOT ADDRESSED BY ERISA The preceding assessment of retirement goals revealed that all workers need additional retirement income beyond Social Security benefits to maintain preretirement standards of living. Yet only half the work force is covered by a supplementary pension plan and IRS integration guidelines allow plans to pay few or no bene- fits to low-paid workers. In addition, private pension benefits are not automatically indexed after retirement, so their real value can decline substantially if inflation accelerates. Although all these

"'Pension Benefit Guaranty Corporation, Premium Requirements for the Single Employer Basic Benefits Insurance Program, Part I Summary Report, table 3, pp. 11-16. 'See ch. 5, table 1. 5*2 Economic Report of the President, January 1984, table B-37, p. 263. According to data published by the Bureau of the Census, 28 percent of single women 65 years old and over fell below the poverty level in 1983. See U.S. Bureau of the Census, Current Population Reports, series P-60, No. 138, Characteristics of the Population Below the Poverty Level: 1983, Advanced Report (GPO, 1984). failures are inconsistent with a sensible retirement income policy, none were addressed by ERISA.

(A) COVERAGE While ERISA attempted to correct perceived abuses and inequi- ties in the private pension system, the legislation did not mandate the establishment of such plans. Consequently, the formation and maintenance of pension plans by employers or through collective bargaining remains a strictly voluntary decision. As a result, pen- sion coverage is far from universal. Statistics on the extent of coverage are subject to considerable controversy; some experts argue that very young, part-time, or mobile workers would not be expected to participate in a compa- ny's pension plan and thus should not be included in coverage sta- tistics. Because of the controversy, two surveys were conducted in 1979 to determine the extent of pension plan coverage. The Presi- dent's Commission on Pension Policy found that 48.1 percent of all active workers eighteen and older were participants in some type of employment-based pension, profit-sharing, or retirement plan in their current jobs.53 Nearly identical results were found in a simi- lar survey of private pension plans conducted by the Census Bureau for the Department of Labor and the Social Security Ad- ministration. 54 Analysis of that survey, however, indicates that the coverage picture improves as the definition of the eligible popula- tion is narrowed. For full-time private sector workers, coverage rose to 50 percent; for those full-time workers over 25 with 1 or more years of service with their current employer, coverage in- creased to 61 percent.55 Because of the influence of industry structure on pension cover- age, some analysts conclude the percentage of the work force cov- ered by pension plans is not likely to increase significantly in the future. Industries with traditionally high pension coverage, such as manufacturing, are expected to employ a declining share of work- ers, while employment in industries with low pension coverage, such as retail trade and services, is projected to increase. Moreover, small businesses, which employ the bulk of noncovered workers, are unlikely to adopt pension plans. These businesses operate on a very tight profit margin in a highly competitive environment and cannot afford the additional cost of a pension plan, especially since the relative cost of establishing such a plan tends to be higher for small firms. Indeed, preliminary tabulations from a May 1983 survey indicate that pension coverage has actually declined in the 1979-83 period.

"President's Commission on Pension Policy, "Preliminary Findings of a Nationwide Survey on Retirement Issues" (May 1980), p. 2. 5Daniel J. Beller, Patterns of Worker Coverage by Private Pension Plans, Department of Labor, Labor-Management Services Administration, Pension and Welfare Benefit Programs (GPO, 1980). For a summary of survey results, see Beller, "Coverage Patterns of Full-Time Em- ployees under Private Retirement Plans," and the companion article by Gayle Thompson Rogers, "Vesting of Private Pension Benefits in 1979 and Change from 1972," Social Security Bulletin, vol. 44 (July 1981), pp. 3-11, 47, and 12-29, respectively. "This figure must be interpreted cautiously because only 58 percent of the private sector workforce fell into the category of full-time, over 25, more than 1 year of service with current employer, and over 1,000 hours of work a year. President's Commission on Pension Policy, "Pre- liminary Findings of a Nationwide Survey on Retirement Issues," p. 3. 198 Again, the precise percentage of the work force covered depends on how the eligible population is defined. Including Government em- ployees, most of whom have pension coverage, the percentage of the work force covered by pension plans has fallen from 61.1 to 56.2 percent. In terms of the private sector only, coverage has fallen from 51 to 47.1 percent of private employment or from 56.2 to 51.1 percent of private wage and salary workers. Regardless of the coverage figure selected, the fact remains that a large part of the U.S. work force is not covered by a private pen- sion plan. In most industries the employees who are not covered are heavily concentrated in small companies that tend to be at the lower end of the wage scale.5 6 The concentration of coverage among higher paid workers is shown in table 8. Even though this table describes coverage by public as well as private pension plans, it nevertheless shows the strong correlation between income class and pension plan coverage.

TABLE 8.-PERCENT OF NONAGRICULTURAL WAGE AND SALARY WORKERS COVERED BY PENSION PLANS, 1979 AND 1983 5

Percentcovered bypension Earninglevels plan 1983 1979

Total...... 56.2 6 1.1 $1 to $4,999 ...... 24.3 NA $5,000 to $9,999...... 37.5 ...... $10,000 to $14,999...... 57.9 ...... $15,000 toto.$9,9...... $19,999 ...... 71.9...... $20,000 to $24,999...... 79.3 ...... $25,000 to $29,999...... 79.1 ...... $30,000 toto.4999...... $49,999 ...... 83.9...... $50,000 andover...... 84.9 ...... Notreported...... 27.2 ...... Addenda: Private sectoronly: Civilian employment ...... 47.1 51.0 Wageand salary workers ...... 51.1 56.2 NA-not available. , Datafrom Employee Benefit Research Institute/Department of Healthand Human Services, Current Population Survey Pension Supplemental (May1983). Source:Employee Benefit Research Institute, "New Survey Findings on PensionCoverage and Benefit Entitlement," EBRI Issue Brief No. 33 (August1984), tables 1, 8, and9, pp.5, 16and 17. In passing ERISA, Congress recognized that a large segment of the working population was ineligible for the tax advantages asso- ciated with private pension plans. Instead of mandating coverage, however, ERISA established individual retirement accounts for workers who were not participating in a plan and liberalized con- tribution limits on Keogh plans for the self-employed. But while the ERISA provisions broadened the group eligible for tax-deferred

56 Coverage has been shown to be directly related to both wage rates and firm size. In 1972 only 18 percent of private nonfarm workers earning less than $3 an hour were in firms with retirement plan expenditures. The comparable figure for those earning $7 or more an hour was 88 percent. Similarly, only 38 percent of firms with under 100 employees provided retirement benefits as against 93 percent of firms with 500 or more employees. See Donald R. Bell, "Preva- lence of Private Retirement Plans," Monthly Labor Review, vol. 98 (October 1975), p. 18. retirement plans, most of the benefits accrued to higher income persons. Thus, the problem of inadequate coverage remains. Al- though the substantial tax concessions offered higher paid workers have encouraged the growth of private plans, these concessions have not led to widespread coverage among lower paid workers.

(B) INTEGRATION The major conclusion that emerges from a careful consideration of replacement rates provided by Social Security is that low income as well as high income individuals need supplementary retirement income to maintain their preretirement living standards. Hence, little justification exists for allowing plans to "fully" or "maximal- ly" integrate with Social Security so that lower paid workers re- ceive no benefits at all from their pension plans. ERISA failed to address the gaps in protection that are created by the current inte- gration provisions and, therefore, did not go far enough to be con- sidered totally consistent with existing retirement income policy.

(C) INFLATION INDEXING Recommendations from the White House Conferences on Aging have stressed the importance not only of establishing initial bene- fits at appropriate levels, but also of maintaining the value of those benefits over the entire retirement period. Although Social Securi- ty and Federal government pension plans provide automatic cost- of-living adjustments, most private plans do not offer similar pro- tection. Sponsors of private plans, nevertheless, have been aware of the erosive power of inflation and have usually provided some ad hoc adjustments for beneficiaries. The 1980 Bankers Trust study indi- cates that during the 1975-79 period roughly 70 percent of the 325 plans surveyed extended a cost-of-living adjustment to some or all of their beneficiaries. 5 7 However, nearly three out of five plans that provided adjustments did so only once during the period. Ac- cording to Bankers Trust calculations, the average cost-of-living ad- justment over the 1975-79 period for persons who retired in Janu- ary 1975 was equivalent to a one-time increase in their pension of $480, or 9 percent, under a conventional plan and $660, or 8 per- cent, under a pattern plan.5 8 Since the CPI rose 47 percent over the same period, most retirees experienced a substantial decline in their living standards.59 A slightly more favorable picture emerges from a survey of post- retirement cost-of-living adjustments between 1973 and 1979 re- ceived by a nationally representative sample of persons in defined

5" Mischo, Chang & Kaston, Corporate Pension Plan Study, pp. 5, 7. 58 Under conventional plans, which account for % of the bankers trust sample, most firms increased pensions by a stated percentage for each year of retirement. The most common figure was 2 or 3 percent, so that a person who retired in 1975 would have received an 8- to 12percent increase in his pension over the 1975-79 period. Other firms increased pensions by a tlat per- centage, usually 10 percent. And some other firms simply increased benefits by a fixed percent- age that depended on the employee's retirement date; for example, 15 percent if he retired before 1966, 10 percent if he retired between 1966 and 1970, and 5 percent if he retired between 1971 and 1975. Under pattern plans an increase of a flat dollar amount or dollar increase for each year of service was most common. Ibid., pp. 52-55. 69 The CPI rose from 155.4 in December 1974 to 169.3 in December 1979. benefit plans who had retired prior to 1973.60 The mean benefit for this group rose from $2,128 in 1973 to $2,639 in 1979, an increase of 24 percent. The benefit increases were fairly widespread, with 75 percent of the group receiving at least one adjustment. Neverthe- less, since the CPI rose by 63 percent between 1973 and 1979, the cost-of-living adjustments compensated for less than two-fifths of the total price increase. The lack of automatic cost-of-living adjustments means that re- tirees cannot be certain that their private pension benefit will retain its value over their full period of retirement. This type of uncertainty is not consistent with the nation's retirement income goals as expressed by the White House Conferences on Aging and other groups.

4. SUMMARY ERISA generally strengthened the retirement income system in this country, since it appears to have been successful in its attempt to ensure that more people receive larger benefits from their em- ployer-sponsored pension plan. The termination of some defined benefit plans in the wake of ERISA is more accurately attributable to cyclical considerations than to the effect of the legislation. The slowing in the rate of net new plan formation is the inevitable result of industry shifts, and thus would have occurred to some extent in the absence of the legislation. ERISA should not be faulted for what it did not do, since the act was a massive piece of legislation that substantially ameliorated many of the inequities in the private pension system. Nevertheless, several aspects of pensions, such as partial coverage, liberal inte- gration guidelines, and lack of postretirement indexing, which were not addressed by ERISA, leave serious gaps in the Nation's retire- ment system. E. ERISA AND CAPITAL FORMATION Intense interest in capital formation as a national goal coincided with the passage of ERISA. A widely publicized 1974 study by the New York Stock Exchange projected a "capital gap" of $650 billion over the 1974-85 period. 6 1 Another projection, presented by the chairman of the General Electric Co. before the Joint Economic Committee, forecasted a "capital deficiency" of $200 billion over the same period. 62 The entire post-ERISA era has been filled with forlorn comparisons between the saving rate in the United States and that of other countries-particularly Japan-and with at- tempts to increase the U.S. rate of saving through generous tax concessions. Capital formation was not an issue, however, during the deliberations on ERISA and the provisions of ERISA were not

60 Robert L. Clark, Steven G. Allen and Daniel A. Sumner, "Inflation and Pension Benefits," final report under U.S. Department of Labor Contract No. J-9-P-1-0074 (August 1983); and Steven G. Allen, Robert L. Clark and Daniel A. Sumner, "Post-Retirement Adjustments of Pen- sion Benefits," working paper No. 1364 (National Bureau of Economic Research, June 1984). 51 New York Stock Exchange, "The Capital Needs and Savings Potential of the U.S. Economy: Projections through 1985" (NYSE, 1974; processed). 62 See testimony of Reginald H. Jones in Long-term Economic Growth, Hearings before the Subcommittee on Economic Growth of the Joint Economic Committee, 93:2 (GPO, 1974), pp. 66- 77. designed with an eye toward stimulating saving and investment. Nevertheless, ERISA's funding requirements could effect the aggre- gate amount of saving, while its prudent man rule, diversification requirements, and other provisions could effect the composition of financial investments. 1. ERISA AND AGGREGATE SAVING Although it is often assumed that the growth in private pension assets, from $2.4 billion at the end of 1940 to nearly $700 billion in 1984, represents a net increase in saving, economic theory suggests that it may simply reflect a shift in the form of saving. This as- sumption rests on the traditional life-cycle saving model, according to which people try to allocate their resources so as to maintain a steady stream of consumption over their lifetimes." During their working years they do not consume all their income but rather set aside positive saving, which they later dissave in retirement. The life-cycle model predicts that, in an ideal world characterized by perfect labor and capital markets, no taxes, and no uncertainty, people would view pension contributions as a substitute for their own saving. Hence, a worker who received $1 of compensation in pension benefits rather than current wages would be expected to reduce saving in other forms by $1 in order to maintain his initial consumption path. If each worker were to reduce his saving by a dollar, the effect of pensions on saving would be neutral, since the reduction in individual saving would be offset by the increase in pension reserves. Of course, the U.S. economy deviates substantially from the model described above, and these deviations introduce some ambi- guity about the probable effect of private pensions on saving. Fa- vorable tax provisions, imperfect capital markets, and induced re- tirement may cause pension plans to result in an increase in saving. Uncertainty about benefit receipt and amount may either increase or reduce saving, depending on whether people overesti- mate or underestimate their future pension receipts. The fact that pensions are paid as annuities and that private plans are less than fully funded should mean there is less aggregate saving than under the simple life-cycle model. The issue is what do these theoretical possibilities imply about the impact of ERISA on the aggregate amount of capital formation.

(A) THE IMPACT OF FUNDING The provision most likely to affect aggregate saving is ERISA's requirement that existing unfunded liabilities be funded in level annual payments over a period of not more than 40 years and that past-service liabilities created under a newly established plan or by plan amendments be amortized over 30 years. To the extent that plans were previously underfunded, these requirements have un- doubtedly increased the assets held in pension funds. The magni- tude of the impact of these funding requirements on aggregate cap- ital formation, however, is less clear. It depends on the extent to

6 Albert Ando and Franco Modigliani, "The Life-Cycle Hypothesis of Saving: Aggregate Impli- cations and Tests," American Economic Review, vol. 53 (March 1963), pp. 55-84. which underfunding existed prior to ERISA and the degree to which individuals engage in offsetting behavior. With regard to the first point, the IRS had implicitly established minimum funding standards in the pre-ERISA era, requiring firms to contribute annually the normal costs of their plan plus interest accruing on any existing unfunded liability. Recognizing that these unfunded liabilities could pose financial problems in the future, many large plans went beyond the IRS minimum guidelines and made annual payments designed to fund their liabilities over a 30 to 40 year period. Since such funding practices were not uncommon before ERISA, the legislation's mandate has not resulted in a sig- nificant increase in the flow of resources into pension funds. In ad- dition, plans that have had to alter their financing patterns as a result of ERISA are amortizing their liabilities over the next 40 years, thereby introducing additional capital into the market very gradually. Second, the saving behavior of the shareholders of firms with un- funded liabilities may have already offset the negative effect that these plans could have on capital accumulation and will probably adjust to compensate for increased funding. Studies have shown that each dollar per share of a firm's unfunded liability reduces the share price by approximately $1.64 As the share price declines, shareholders' wealth will also decline; they would therefore be ex- pected to reduce consumption and increase saving. Increased saving by shareholders could offset the lack of funding in the pen- sion plan. Similarly, once the firm began to reduce its unfunded li- ability, the process would reverse itself: share prices would rise, shareholders' wealth would increase, and shareholders would in- crease their consumption and reduce their saving. Thus, the overall degree of funding in the pension arena would be expected to have little or no impact on aggregate saving.

(B) REDUCING UNCERTAINTY Another factor that could affect the level of aggregate saving is the extent to which individuals reduce their own saving in re- sponse to saving through a pension plan. Most studies show that, as predicted under the life-cycle model, consumers lower their saving in other forms to compensate for the promise of pension benefits. 65 Although there is general agreement that the reduction in other saving is less than dollar for dollar, to guess about the magnitude of this offestting effect is hazardous in light of the scant empirical evidence. Nevertheless, a reasonable conjecture might be that workers lower their nonpension saving by 65 cents for each $1

"George S. Oldfield, "Financial Aspects of the Private Pension System," The Journal of Money, Credit and Banking, vol. 9, pt. 1 (February 1977), pp. 48-54; Martin S. Feldstein, "Do Private Pensions Increase National Saving?" Discussion Paper 553 (Harvard Institute of Eco- nomic Research, 1977), pp. 20-21; Mark Gersovitch, "Economic Consequences of Unfunded Vested Pension Benefits, ' Working Paper 480 (National Bureau of Economic Research, 1980); Martin Feldstein and Stephanie Seligman, "Pension Funding, Share Prices and National Saving," Working Paper 509 (National Bureau of Economic Research, 1980). 6 See for example, Peter Diamond and Jerry Hausman, "Individual Savings Behavior," paper prepared for the National Commission on Social Security (May 1980); Martin S. Feldstein and Anthony J. Pellechio, "Social Security and Household Wealth Accumulation: New Microeco- nomic Evidence," Review of Economics and Statistics, vol. 61 (August 1979), pp. 361-368. Alicia H. Munnell, "Private Pensions and Saving: New Evidence," Journal of Political Economy, vol. 84 (October 1976), pp. 1013-1032. of saving through pension plans, resulting in a net increse in ag- gregate saving of roughly 35 cents. ERISA could have increased the amount of offsetting saving be- havior by reducing the uncertainty about the receipt of future ben- efits. With vesting standards clarified and benefits insured by the PBGC, workers covered by a pension plan can be considerably more confident of ultimately receiving retirement benefits. This in- creased certainty would allow workers to cut back further on their nonpension saving. Since no evidence exists to support the conten- tion that any such change in behavior has occurred, however, the most reasonable conclusion is that ERISA has probably had little impact on the level of capital accumulation in this country.

2. ERISA AND THE COMPOSITION OF INVESTMENT Several provisions of ERISA might be expected to increase con- servatism in investment decisions and to encourage investment in assets other than common stock. ERISA mandates "fiduciary responsibility" for pension plan trustees, for investment managers, and for any other person who may have control over the pension plan. The act states that fidu- ciaries must manage pension assets for the sole interest of the par- ticipants and beneficiaries and that they must invest with the skill and diligence of a "prudent man." More specifically, fiduciaries cannot use plan assets for their own benefit; allow business and in- vestment transactions between the plan and "parties-in-interest," such as the sponsoring employer, plan participants, investment managers, or the unions involved; or invest more than 10 percent of the plan's assets in securities of the employer. The act also re- quires that the fiduciary diversify the pension portfolio in order to minimize investment risk. A fiduciary is personally liable to the plan for losses resulting from violations of his fiduciary obligations and, under certain circumstances, those of cotrustees. Moreover, ERISA authorizes civil penalties for parties-in-interest who engage in prohibited transactions. These provisions should discourage plan trustees from taking any undue risk with plan assets. The bias against risky assets is reinforced by funding require- ments that make large capital losses undesirable. The law requires companies to reassess their actuarial assumptions in light of invest- ment gains and losses every 3 years. When a plan's actual results deviate significantly from those projected by its actuaries, the ad- ministrators of the plan must adjust contributions according to "ex- perience" gains or losses and amortize the investment gains or losses over 15 years. Although the impact of these adjustments on pension costs is likely to be considerably less than changes in bene- fit or salary assumptions, the possibility of increased fluctuations in annual contributions would be expected to lessen the attractive- ness of risky investments. A final factor contributing to the adoption of more conservative investment policies is the contingent liability created by the PBGC insurance provisions. Before 1974, firms were not legally liable for the payment of unfunded vested pension benefits. If a firm termi- nated its pension plan, employees with vested pension claims would receive benefits only insofar as the pension fund was adequate. 204 Such a lack of legal responsibility can lead to an aggressive invest- ment posture, since, in the case of an insufficiently funded pension plan, the sponsor could profit significantly from a high risk-high return strategy. If the investment were successful, the sponsor gained; if the investment failed, the claimants could conceivably forfeit part or all of their benefits. The PBGC now guarantees the payment of "basic benefits," which are roughly equivalent to vested benefits under the particu- lar plan. ERISA imposes a limit on the amount insured, which has increased from $750 per month in 1974 to $1,602 per month in 1984. When a plan terminates without sufficient assets, the PBGC becomes the trustee of the plan and can hold the sponsoring com- pany liable for the underfunding up to an amount equal to 30 per- cent of the company's net worth. Hence, ERISA has altered the dis- tribution of risk associated with portfolio performance. Vested workers under the established plans are now assured of receiving benefits up to the guaranteed level and the liability for their guar- antee is borne in large part by the sponsoring firm.6 6 This in- creased accountability on the part of the plan sponsor could be ex- pected to produce a more conservative investment strategy. Despite the potential for ERISA to cause pension fund managers to engage in less risky investment behavior, no major shifts in in- vestment patterns have occurred since its passage. As shown in table 9, the distribution of assets held in pension trusts have re- mained relatively unchanged over the past decade; the only notice- able development being that the holdings of Government securities have increased at the expense of corporate bonds. Because life in- surance pension assets are combined with other assets managed by life insurance companies, their portfolio mix is generally assumed to be similar to that of the overall holdings of the life insurance industry. Based on this assumption, again no striking changes in investment patterns emerge. Combining the assets held by life in- surers and those held in pension trusts reveals a very stable portfo- lio distribution for private pension plans over the last decade. TABLE 9.-DISTRIBUTION OF PRIVATE PENSION ASSETS HELD INPENSION TRUSTS AND WITH LIFE INSURANCE COMPANIES, 1974 AND 1984 1

Heldin pensiontrusts Heldwith life Total Assets insurers1 1974 1984 1974 1984 1974 1984

Amount (billions) ...... $115.5 $408.8 $60.8 $276.0 $176.3 $684.8 Percentagedistribution: Corporateequities...... 54.8 59.4 9.6 11.4 39.2 40.1 Corporatebonds 4...... 29.4 16.7 42.3 39.6 33.8 25.9 U.S.Government securities...... 4.8 18.4 1.9 10.1 3.8 15.0 Mortgages...... 2.1 1.3 37.8 27.0 14.4 11.6

66 For a detailed discussion of how the risk is apportioned among the firm, the participants, and the PBGC before and after ERISA for plans with different initial funding positions, see Munnell, The Econmics of Private Pensions, pp. 145-46; also Jeremy Bulow, "Analysis of Pen- sion Funding Under ERISA," Quarterly Journal of Economics, vol. 97, no. 3 (August 1982), pp. 435-452. Cashand deposits...... 4.3 2.3 .8 .7 3.1 1.7 Other assets...... 4.6 1.8 7.5 11.4 5.6 5.7

Firstquarter 1984. 2 Includesfunds held in specialaccounts. Figuresmay not add due to rounding. Includesforeign hoods. Sources:Board of Governorsof the Federal Reserve System, FRow of FundsAccounts, Assets and Liabilities Outstanding 1959-82 (Washington, DC,1983). p. 24 andSector Statements of Financial Assets and Liabilities (July 1984). Supporting the contention that ERISA caused little change in the investment patterns of private pension plans are the results of a detailed study, by three professors at the Wharton School at The University of Pennsylvania, of investment practices before and after ERISA. 67 The analysis concludes that while significant shifts occurred in the asset demands of private pension plans after the passage of ERISA, the portfolio restructuring reflected market fac- tors and is not attributable to the legislation. Hence, ERISA ap- pears to have had little impact on the distribution of assets within private pension portfolios and hence on the overall composition of investment in the U.S. economy.

3. SUMMARY Although increasing capital formation is currently an important national goal, it was not a major priority during the evolution of ERISA. The funding, fiduciary, and insurance provisions in ERISA were intended to secure the rights of plan participants, not to affect the level or composition of investment. Because to a large extent the law merely codified existing practices in this area, ERISA has probably had little impact on either the aggregate amount of capital accumulation or on the pattern of investment of accumulated assets. F. CONCLUSION ERISA was an important and successful piece of legislation that ameliorated many of the problems that existed in the private pen- sion system. No longer do workers forfeit pension rights because of overstrict participation and vesting requirements, mismanagement of pension assets, or the termination of insufficiently funded plans. As a result of the legislation, more people covered by qualified pen- sion plans end up receiving benefits. The major goal of ERISA, to ensure the receipt of benefits, is fully consistent with the objectives of federal tax policy. The favor- able tax provisions associated with qualified pension plans have been designed to induce higher paid employees to adopt retirement savings plans that will also benefit the rank and file. The provi- sions of ERISA were aimed at ensuring that the rank and file re- ceive their promised benefits. The only exception in ERISA to gen- erally accepted tax policy goals was the introduction of IRA's. While these accounts were developed in an attempt to expand the opportunity for tax-deferred saving, their lack of nondiscrimination requirements has resulted in considerably greater usage among high-income as compared to low-income workers. The consistency of ERISA with national employment goals is dif- ficult to assess, since the interests of individual employers and

87 J. David Cummins, John Percival, and Randolph Westerfield, The Impact of ERISA on the Investment Policies of Private Pension Funds and Capital Market Efficiency, final report under U.S. Department of Labor Contract No. J-P-P-6-0209 (April 1980). public policymakers diverge dramatically. For example, ERISA has probably encouraged earlier retirement by ensuring that more workers receive private pension benefits to supplement Social Secu- rity. This outcome may be consistent with the preferences of em- ployers, who established pension plans in large part to retire super- annuated workers, but it is clearly not consistent with public ef- forts to lengthen the average work life and postpone the normal re- tirement age. Similarly, more rapid vesting has probably increased mobility among workers with more than 10 years of service. Great- er mobility may be desirable from the perspective of economic effi- ciency, but the interest of the individual employer rests with re- taining highly trained workers. The main thrust of ERISA-namely, to ensure the receipt of pri- vate pension benefits-is also fully consistent with national retire- ment income goals. Since Social Security alone is not sufficient to maintain preretirement living standards at any income level, most individuals would experience a decline in their economic well-being upon retirement without supplementary private pension benefits. By ensuring that more workers covered under a private pension plan received the benefits to which they are entitled, ERISA has made progress toward the President's Commission on Pension Poli- cy's recommended goal of fully replacing preretirement disposable income. Finally, in the area of capital formation, ERISA's provisions-in particular, the institution of funding standards-are consistent with national efforts to increase saving and capital formation. However, because many plans were fairly well funded prior to ERISA and because offsetting behavior by individuals may have mitigated any adverse effect of this underfunding on aggregate saving, the net impact of ERISA on saving and capital has prob- ably not been very significant. Since most of ERISA's major provisions seem consistent with na- tional goals, with the possible exception of the introduction of IRAs and the effect of increased benefit security on retirement behavior, the reader may wonder about the source of ERISA's numerous al- leged shortcomings discussed throughout the body of the chapter. Most of the areas where inconsistencies were found between na- tional goals and present pension policy arise from issues not ad- dressed by ERISA. For example, in the area of tax policy, ERISA failed to consider ways to ameliorate inequities created by the cur- rent integration guidelines or to explore alternatives to a totally employer-financed pension system. In employment policy, ERISA failed to address the erosive impact of inflation on the vested bene- fits of mobile employees and did not consider ways to reduce the higher costs of employing older workers. The greatest failings of ERISA appear in the area of retirement income policy, where par- tial coverage, liberal integration guidelines, and lack of postretire- ment indexing leave serious gaps in the Nation's retirement system. The major conclusion that emerges from the preceding analysis, therefore, is that ERISA was a much needed and extremely suc- cessful piece of legislation that ameliorated many of the document- ed inequities in the private pension system. But some significant problems still remain, since Congress was unable to address all in- 207 equities in 1974. In addition, other concerns have become more pressing since the passage of ERISA, such as the effect of inflation on the value of benefits over the work life and in retirement. The continuation of past problems and the development of new con- cerns indicate that, a decade after ERISA, it may be time to begin to formulate new pension legislation. 0